Sharing Lives, Dividing Assets: An Inter-Disciplinary Study 9781472564856, 9781841132594

Citation preview

Preface The idea for this project, and so this book, originated in our growing awareness of two things. First, that the courts’ decisions about appropriate basis for dividing wealth on divorce or allocating ownership of property on separation were being increasingly influenced by how parties (are perceived to) organise their lives while together and by what that was supposed to indicate, for example, about their intentions regarding ownership or about the economic impact of relationship breakdown, but that the judges were not always apprised of all the research data necessary to make a fully informed decision about those issues. Secondly, that there is a vast social science literature touching on many issues relevant to legal decision-making in this area, but that researchers in those other disciplines generally run along parallel tracks that never meet each other—or lawyers. We thought that we could usefully try to make a contribution on both fronts by bringing together people working in a wide range of disciplines pertinent to the broad theme of money, property, relationships and relationship breakdown, in order to stimulate debate and cross-fertilisation of ideas between them, and to bring their valuable research to the attention of a wider audience. The project, run under the aegis of the Cambridge Socio-Legal Group, based at the University of Cambridge, adopted a workshop model. Invited contributors circulated draft chapters on their research in advance of a two-day meeting at which each contributor was then able, following a short presentation, to discuss their work with the whole group. This methodology proved highly successful in enhancing inter-disciplinary understanding, not least in identifying and unravelling areas of misunderstanding, often generated simply by the use of seemingly familiar language in an unfamiliar way. It was also highly productive of ideas for new lines of inquiry in all of the disciplines represented at the workshop, thanks to the fresh perspectives cast on otherwise familiar territory by those whose disciplinary background provided a different vantage point from which to view the material. In light of those discussions, the contributors were then able to revise their chapters before publication, and our introductory chapter benefited considerably from insights offered by members of the group during our discussions over those two days. We hope that the result is a volume that provides a valuable resource, giving access to findings on wide range of issues from a wide range of disciplines, in a way that will both inform legal development in this field and encourage further research. We are extremely grateful to the Cambridge Socio-Legal Group, the John Hall Fund of the University of Cambridge, and to Trinity College, Cambridge for their financial support for this project, and to the staff of Trinity College who helped the workshop to run so smoothly. In addition to the contributors themselves,

vi Preface whose biographies appear in the following pages, we must also acknowledge a huge debt to the following individuals for their contributions to discussion during the workshop: Belinda Brooks-Gordon, Ceridwen Roberts, Martin Richards, Brian Sloan and Anke Zimmermann. Finally, we must also thank Liam D’Arcy Brown for his assistance with proof-reading and editing.

Notes on Contributors Anne Barlow studied Law with French at the University of Sussex and is Professor of Family Law and Policy at the University of Exeter. The main focus of her research has been on family and housing law, and especially cohabitation law, and she has published widely in these fields. In particular, she is author of Cohabitants and the Law (Butterworths, 2001) and (with S Duncan, G James, and A Park) of Cohabitation, Marriage and the Law (Hart Publishing, 2005). She has, since 2000, directed and co-directed a number of empirical socio-legal projects funded by the Nuffield Foundation, the Ministry of Justice and the Economic and Social Research Council (ESRC) including ‘Family Restructuring, the Common Law Marriage Myth and the Need for Legal Realism, Community of Property—A Regime for England and Wales?’ and ‘The Common Law Marriage Myth and Cohabitation Law Reform Revisited’. She has been consulted by the Law Commission for England and Wales, the Scottish Executive, and the Office of Law Reform, Northern Ireland on reform of cohabitation law and the implications of her empirical research for this process. Carole Burgoyne is Senior Lecturer in the School of Psychology, University of Exeter. Her main research interests are in economic psychology and include studies of money, marriage, currency change and distributive justice. Her most recent projects are ‘The Common Law Marriage Myth Revisited’ (led by Anne Barlow); ‘Financial Management Practices in Non-traditional Heterosexual Couples’ (with Stefanie Sonnenberg and Anne Barlow); ‘Beliefs about the Allocation of Money in Marriage’ (with Stefanie Sonnenberg); and ‘Money management in Lesbian and Gay Couples’ (with Victoria Clarke). Shirley Dex is Professor of Longitudinal Social Research in the Centre for Longitudinal Studies, Institute of Education, University of London. She has held posts at the Judge Business School, University of Cambridge, the Institute for Social and Economic Research at the University of Essex, and the Economics Department at the University of Keele. From 1998 to 2003 she was Research Advisor to the Joseph Rowntree Foundation’s Work and Family Life Programme. She has published widely on women’s employment and cross-national comparative research, equal opportunities, families and work, ethnic minorities and employment, flexible working arrangements in organisations, work and care, and family policy. Recent books include Families and Work in the 21st Century (Joseph Rowntree Foundation, 2003). She has also co-edited (with Heather Joshi) Children of the 21st Century (The Policy Press, 2005) and (with Jacqueline Scott and Heather Joshi) Women and Employment: Changing Lives and New Challenges (Edward Elgar Publishing Ltd, 2008).

x Notes on Contributors Antony Dnes currently holds a chair in Economics at the University of Hull, and has held previous appointments at the Universities of Birmingham, Edinburgh and St Andrews in the UK, and at the Virginia Tech and the George Mason Law School in the US. He completed his PhD in Economics at the University of Edinburgh, and has an LLB from the University of London. He is a specialist in the economic analysis of law, with a number of published articles dealing with family topics, in both law and economics journals. He co-edited The Law and Economics of Marriage and Divorce (CUP, 2002), a collection of articles from leading academics on both sides of the Atlantic. Gillian Douglas obtained her LLB at the University of Manchester in 1977 and her LLM at the London School of Economics in 1978. She is currently Professor of Law and Head of Cardiff Law School, having previously taught at the University of Bristol and the National University of Singapore. She is the co-editor of the Child and Family Law Quarterly and of the Case Reports section of Family Law. Her publications include (with Nigel Lowe) Bromley’s Family Law (9th and 10th editions) and An Introduction to Family Law (Clarendon Law Series, 1st and 2nd editions). With colleagues from disciplines including psychology and social work she has conducted a number of empirical studies in law, including How Parents Cope Financially on Marriage Breakdown (2000, funded by the Joseph Rowntree Foundation), Grandparenting in Divorced Families (2004, funded by the Nuffield Foundation), and A Failure of Trust: Resolving Property Disputes on Cohabitation Breakdown (2007, funded by the ESRC). Sally Dowding is a District Judge based at Birmingham Civil Justice Centre. She obtained an LLB degree from the University of Manchester in 1973 and subsequently qualified as a solicitor. For most of a practising career spanning nearly 30 years she specialised in family law. She was a member of the Law Society’s Family Law Committee and Children Law Sub-committee for the 10 years preceding her judicial appointment and chaired the latter committee for nearly four years. John Eekelaar is Emeritus Fellow of Pembroke College, Oxford, and a Fellow of the British Academy. He has written and conducted research in family law since the 1970s, and is currently Joint Director of the Oxford Centre for Family Law and Policy with Mavis Maclean, with whom he has co-authored several books. His latest book, Family Law and Personal Life, was published by Oxford University Press in hardback in 2006, and paperback in 2007. He also co-edits the International Journal of Law, Policy and the Family. Andrea Finney is a Research Fellow in the Personal Finance Research Centre at the University of Bristol. Primarily a quantitative researcher, she previously spent six years at the Home Office and, latterly, the Office for National Statistics, where she was involved in rolling-out the new longitudinal survey of household assets. Andrea’s current research interests lie in furthering understanding of saving and borrowing behaviour, financial vulnerability and decision-making. She is co-author of reports on over-indebtedness (for the European Commission),

Notes on Contributors xi the results from the baseline survey of consumer purchasing (for the Financial Services Authority) and modelling exposure to financial risk (for Genworth Financial). Hayley Fisher is a PhD student in the Faculty of Economics at the University of Cambridge, supervised by Thomas F Crossley and Hamish Low. Her research interests include family and labour economics, and she is currently considering the economic impact of divorce legislation on the decision to marry or to cohabit. Emma Hitchings is Senior Lecturer in the School of Law, University of Bristol. She studied law as an undergraduate before taking her PhD in 2003. Her main research interests lie in the field of family law, where she has written and worked on issues concerning domestic violence, adoption, ancillary relief, and the recognition of adult relationships, particularly same-sex couples. She is case commentaries editor for the Journal of Social Welfare and Family Law and an academic member on the Family Law Committee of the Law Society of England and Wales. She is currently engaged in research on pre-nuptial agreements for the Law Commission for England and Wales. Jane Lewis is Director of Research, Evidence and Evaluation at the National Children’s Bureau and was formerly Director of the Qualitative Research Unit at the National Centre for Social Research. She practised as a solicitor before moving into social policy research. Her recent research includes Separating from Cohabitation: Making Arrangements for Finances and Parenting (2006, funded by the Department for Constitutional Affairs); Settling Up: Making Financial Arrangements after Divorce or Separation (2002, funded by the Nuffield Foundation), and Pensions and Divorce: A Qualitative Study among Solicitors (2000, funded by the Department for Work and Pensions). She has written widely on research methods and is co-editor of Qualitative Research Practice, published by Sage in 2003. She has a long-standing interest in relationship breakdown and more broadly in research on children and young people’s lives; social inclusion, inequality and disadvantage; and research methods. Hamish Low is a Senior Lecturer in the Faculty of Economics at the University of Cambridge and a Fellow of Trinity College, Cambridge. He is also a research fellow at the Institute for Fiscal Studies. His main research interests are in understanding the risks that individuals face over their lifetimes and in analysing the extent to which these risks are insured. A particular interest has been in understanding changing patterns of women’s labour supply across different generations. He has published primarily in economic journals, such as the American Economic Review and the Journal of Public Economics. Mavis Maclean CBE is Joint Director with John Eekelaar of the Oxford Centre for Family Law and Policy, Department of Social Policy and Social Work, University of Oxford. She also acts as Academic Adviser to the Ministry of Justice. Recent

xii Notes on Contributors publications include Parenting after Partnering (Hart Publishing, 2007), Cross Currents in Family Law, edited with Katz and Eekelaar (OUP, 2000), and Family Lawyers with Eekelaar and Beinart (Hart, 2000). She is a former President of the Research Committee for the Sociology of Law and a Fellow of the International Institute for Sociology of Law, Onati, Spain. Joanna Miles is a University Lecturer in the Faculty of Law and Fellow of Trinity College, University of Cambridge. Her research interests lie mainly in the area of family property and financial remedies following relationship breakdown, whether marital or non-marital. Her work has been considered and cited judicially in England and Wales, Ireland, and New Zealand. She was recently seconded to the Law Commission for England and Wales for two years to work as a lawyer on its project ‘Cohabitation: The Financial Consequences of Relationship Breakdown’. She is currently embarking on a Nuffield Foundation-funded project examining the first three years’ operation of the new remedies between cohabitants in Scotland, introduced by the Family Law (Scotland) Act 2006, with Professor Fran Wasoff at the University of Edinburgh. She is co-author (with Sonia Harris-Short) of Family Law: Text, Cases, and Materials (OUP, 2007). Julia Pearce graduated from the University of Bristol in 1988 with a degree in Law, subsequently practising as a solicitor specialising in family law for seven years. She obtained her LLM in 1996, following which she left private practice to work full-time in socio-legal research. Joining the Law Department of Bristol University she has worked on a number of research studies in family law with colleagues in the Schools of Law and of Social Policy. These include A Study of Children Act Applications (1998, funded by the Nuffield Foundation), An Evaluation of the Family Mediation Pilot Scheme (1999, for the Legal Services Commission), A Survey of Ancillary Relief Outcomes (1999, for the Lord Chancellor’s Department), and A Failure of Trust: Resolving Property Disputes on Cohabitation Breakdown (2007, funded by the ESRC). Debora Price is a Lecturer in Social Policy at the Institute of Gerontology at King’s College London, where she specialises in the study of pensions and the poverty of older people, issues relating to household money over the life-course, and the legal regulation of the financial consequences of family formation and dissolution. She also has an interest in survey methodology and data analysis, regularly undertaking consultancy work in this area. Formerly a barrister specialising in family law, her ESRC PhD research focused on the impact of family change on pension scheme participation in the UK. She has recently concluded an ESRC/DWP fellowship placement at the Department for Work and Pensions reviewing measures of pensioner poverty. Her current research includes the ESRC-funded project ‘Behind Closed Doors: Older Couples and the Management of Household Money’. Rebecca Probert is a Senior Lecturer at the University of Warwick, teaching family law, child law, and a course on ‘law and the intact family’. She has written

Notes on Contributors xiii widely on the topics of cohabitation and marriage, particularly from a historical perspective. Her recent publications include Marriage Law and Practice in the Long Eighteenth Century: A Reassessment (CUP, 2009) and ‘Looking back on the overlooked: cohabitants and the law 1857–2007’ in N Lowe and G Douglas (eds) The Continuing Evolution of Family Law (Jordans, 2009). She is also co-author (with J Masson and R Bailey-Harris) of Cretney: Principles of Family Law (Sweet & Maxwell, 2008), editor of Family Life and the Law: Under One Roof (Ashgate, 2007), and co-editor (with Stephen Gilmore and Jonathan Herring) of Responsible Parents and Parental Responsibility (Hart Publishing, 2009). Jacqueline Scott is Professor in Empirical Sociology at the University of Cambridge and a Fellow of Queens’ College. She directs the ESRC’s Research Network on Gender Inequalities in Production and Reproduction. Her former positions include Director of Research at the ESRC Centre for Micro Social Change at the University of Essex, where she was involved in the original design and implementation of the British Household Panel Survey. Her research interests include life-course research; gender-role change; attitudinal research; and ageing and well-being. She has recently co-edited (with Judith Treas and Martin Richards) the Blackwell Companion to the Sociology of Families (Blackwell, 2004); (with Yu Xie) the Sage Benchmark Series on Quantitative Sociology (Sage, 2005); and (with Shirley Dex and Heather Joshi) Women and Employment: Changing Lives and New Challenges (Edward Elgar Publishing Ltd, 2008). Stefanie Sonnenberg is a Lecturer in Psychology at the University of Portsmouth. She graduated from the University of Exeter with an MSc in Economic Psychology (1998) and completed her PhD in Social Psychology at the University of St. Andrews (2004). Her general research interest lies at the intersection between economic psychology and social psychology, and her work primarily focuses on the relationship between identity and economic practices. Her recent projects (with Carole Burgoyne and Anne Barlow) have examined financial management practices in intimate relationships. Jean Taylor is an Associate Fellow at the Office for Public Management (OPM). Prior to joining OPM, Jean was a Senior Researcher at National Centre for Social Research (NatCen), where she undertook fieldwork for, and co-authored, Separating from Cohabitation: Making Arrangements for Finances and Parenting (2006, funded by the Department for Constitutional Affairs). Jean holds an MSc in European Social Policy Analysis from the University of Bath, and a BA in Modern History from the University of Oxford. Rosalind Tennant is a Research Director at NatCen. She specialises in qualitative research and has worked across a range of social policy areas including families and work, older people’s services, and social inclusion. She is co-author of the research report Separating from Cohabitation: Making Arrangements for Finances and Parenting (2006, funded by the Department for Constitutional Affairs). Her current work focuses on the areas of learning, skills, and employment,

xiv Notes on Contributors in particular special educational needs. Rosalind graduated in 2000 with a BA in Communication and in 2001 with a MA in Research in Language and Communication, both from the University of Wales, Cardiff. She worked in qualitative and quantitative market research for a number of years before joining NatCen’s Qualitative Research Department in 2004. Carolyn Vogler is Reader in Sociology at City University, London. She has published extensively in the areas of globalisation, social class, unemployment, gender segregation in the labour market, and money and gender relations within intimate relationships. Currently, her main research interests are in the area of money, power, and inequality in new forms of intimate relationships. Hilary Woodward graduated from the London School of Economics with a degree in Sociology and Social Anthropology in 1967. She worked as a research associate on projects with the Institute of Education, the Institute of Psychiatry and the Medical Research Council, before taking up training in law. She practised as a family lawyer from 1982 to 2003 and was accredited as a Resolution cohabitation specialist in 2001. She has worked as a family mediator with Bristol Family Mediation since 1993, and joined Cardiff Law School as a research associate in 2003. She is co-author of ‘Ancillary Relief Outcomes’ (2000) 12 Child and Family Law Quarterly 43, Divorce for Dummies (John Wiley and Sons Ltd, 1st and 2nd editions) and A Failure of Trust: Resolving Property Disputes on Cohabitation Breakdown (2007, funded by the ESRC).

1 Sharing Lives, Dividing Assets Legal Principles and Real Life JOANNA MILES AND REBECCA PROBERT

INTRODUCTION

A

perennial debate in family law concerns the principles that should be applied in determining the financial and property consequences of relationship breakdown. Key questions include the following: should the principles (whatever they may be) be the same for married1 and cohabiting couples? What principles should be adopted, for both or for either category of relationship? Should the division of resources, and so the applicable principles, reflect the parties’ own expectations or societal norms? If the former, how are those expectations or intentions to be discerned? If the latter, how should policy-makers— whether judicial or legislative—determine what those norms are or should be? The House of Lords has endeavoured to take heed of empirical data in its deliberations on some of these questions in recent cases concerning the financial consequences of divorce2 and the separation of cohabitants.3 The Law Commission similarly drew on socio-legal research4 in devising its recommendations for reform of the law governing cohabitants’ financial remedies following separation.5 More recently, it cited the lack of large-scale empirical research that would provide a contemporary socio-economic background for law reform amongst its reasons for declining to take on a project considering the financial remedies available on the dissolution of marriage and civil partnership.6 The importance of both quantitative and qualitative socio-legal research in informing the development and reform of family law is therefore widely accepted. But if family lawyers and policy-makers 1 The same questions apply to civil partners, although, for the reasons set out below, the focus of this collection is on opposite-sex couples. 2 Miller v Miller, McFarlane v McFarlane [2006] UKHL 24 (referred to as ‘Miller; McFarlane’). 3 Stack v Dowden [2007] UKHL 17. 4 For example research concerning couples’ money-management practices and public attitudes towards law and law reform in this area: Vogler (2005); Burgoyne et al (2006); Vogler, Brockmann and Wiggins (2006, 2008). 5 Law Commission (2006, 2007). 6 Law Commission (2008) [5.6].

4 Joanna Miles and Rebecca Probert are to engage successfully with empirical data in this sphere then, as the Law Commission’s recent decision about its future programme of work indicates, much more research is needed, not least about the arrangements people make during their relationships and about the operation and impact of the current law. This inter-disciplinary volume is intended to make some contribution towards addressing this gap in our knowledge—and to overcome perceived difficulties of accessing material from what might otherwise (to the legal audience, in particular) be an unfamiliar literature—by bringing together leading academics in the fields of law, economics, sociology, psychology and other social sciences. It seeks to provide a critical examination of issues relating to the financial relationships and obligations between couples (both during their relationships and on separation) by collating empirical research conducted from the perspectives of various social science disciplines. Part I of the book, of which this chapter forms a part, sets out the legal framework and identifies the importance of empirical studies for this area. If legal principles are to take account of what couples do in practice, then information about practice is essential. It is therefore crucial to build up a body of research on the ways in which couples behave and, very importantly, what meaning can reliably be attributed to that behaviour in order to inform what will otherwise be ‘instinctive’, but quite possibly incorrect, assessments of these matters. Part II accordingly examines how couples, whether cohabitants or spouses, manage their various resources—monetary, property and human—during their relationships. Part III considers the impact that the law currently has on separating couples, examining how legal principles translate into reality and what their consequences are for the parties. This requires separate consideration of cohabitants and spouses, since significantly different rules currently apply to each group. Finally, Part IV considers the issue of legal rationality: it may be rational for the law to be shaped by patterns of behaviour, but how far will individual couples allow their behaviour to be shaped by the law? The focus on opposite-sex couples is both necessary and justified: necessary, because much7 of the research carried out to date has focused exclusively on married couples and opposite-sex cohabitants, and one cannot assume that it can be transposed to couples of the same sex (and still less to those who are not in couple relationships8); and justified, because one of the most striking findings 7 There is a growing literature on the family practices and perspectives of same-sex couples (see eg Weeks, Heaphy and Donovan (2001); Smart, Mason and Shipman (2006), Clarke, Burgoyne and Burns (2007); Burns, Burgoyne and Clarke (2008); Peel and Harding (2008)). The data is still limited in comparison to that available for opposite-sex couples, inevitably so for civil partnership, given its recent introduction. 8 Although the position of wider kin has attracted more attention in recent months in the wake of Burden and Burden v UK (2008) 47 EHRR 38, and the proposals contained in the consultation paper issued by Resolution and Lord Lester in 2008 (see www.resolution.org.uk/livingtogether/). For literature in this area, see eg the work of Roseneil and others on friendship and non-conventional partnerships within the CAVA research project: www.leeds.ac.uk/cava/research/strand3d.htm; Williams (2004); Bottomley and Wong (2006).

Sharing Lives, Dividing Assets 5 that permeates all of the chapters is the continuing relevance of gender to the division of both responsibilities and wealth, as Scott and Dex discuss in chapter three. This has implications for the way in which money is dealt with during the relationship and the resources available to each party, as illustrated by Vogler (chapter four), and Burgoyne and Sonnenberg (chapter five) in their examinations of money management practices, and by Finney (chapter six), focusing on the issue of debt. Crucially, it also has a profound impact upon relationship breakdown, both in the immediate aftermath, as Lewis, Tennant and Taylor show in their discussion of disadvantage on the breakdown of cohabiting relationships in chapter eight; and in the longer term—quite how long being revealed by Fisher and Low, in their analysis of post-separation income levels in chapter eleven, and by Price, in her discussion of pension assets in later life in chapter twelve. THE LEGAL AND POLICY CONTEXT

The laws currently applicable to financial disputes that may arise between spouses and cohabitants are complex and, in key respects, different. During the currency of these relationships, much of the law applies even-handedly to both types of relationship. While spouses enjoy certain limited statutory entitlements (notably statutory rights to occupy the shared home9), they are for the most part treated in private law10 as entirely separate individuals, each equally entitled to earn, acquire and own property, and to manage those resources independently of the other party.11 Indeed, they are formally treated no differently from cohabitants. The same rules of property and trust law apply across the board in order to determine who owns what—in particular, who owns the family home and in what shares (if any). The legal position of spouses and cohabitants diverges substantially, however, at the point of relationship breakdown. While both may apply for the courtbased remedy of tenancy transfer,12 only spouses have access to a more general jurisdiction which enables the discretionary redistribution of income and capital resources, present and future.13 Cohabitants, by contrast, are left with whatever the law of property and trust entitles them to, supplemented (or reduced) only by

9 Family Law Act 1996 ss 30–34. See also the statutory duty to maintain during marriage (in practice substantially less important than maintenance following divorce): eg Matrimonial Causes Act 1973 s 27. Other spousal rights and duties are of rather less practical relevance in the current social context, and have not been extended to civil partners: Married Women’s Property Act 1964 s 1 (determining the ownership of savings made by the wife (specifically) from a housekeeping allowance); the ‘presumption of advancement’ whereby transfers of wealth made from husband to wife (but not vice versa) are presumed (now only weakly—Pettitt v Pettitt [1970] AC 777) to be intended as a gift. See generally Harris-Short and Miles (2007) ch 3; Masson, Bailey-Harris and Probert (2008) chs 3 and 4. 10 Contrast the aggregation of spouses’ (and cohabitants’) resources for the purpose of calculating entitlement to means-tested benefits and tax credits. 11 Since the Married Women’s Property Acts of the late 19th century. 12 Family Law Act 1996 s 53 and Sch 7. 13 Matrimonial Causes Act 1973 Part II.

6 Joanna Miles and Rebecca Probert whatever is afforded (or demanded) by the law relating to the financial support of children.14 We outline the details of these laws further below. It is likely that many people are surprised by these key features of the current law. For example, we know from the British Social Attitudes survey that a majority (albeit shrinking) still positively believe (in general terms15) that spouses and cohabitants enjoy similar legal status, leaving many cohabitants disappointed (and seriously disadvantaged) on relationship breakdown.16 Conversely, a large number of spouses would probably be surprised to discover that they do not automatically—by virtue of marriage—enjoy joint ownership of all or certain assets (particularly the matrimonial home and/or its essential household contents), as is the case in certain European jurisdictions operating systems of ‘community of property’.17

The Law of ‘Ancillary Relief ’ on Divorce Upon divorce, the court has a discretion as to how it should exercise its wideranging powers to reallocate assets between the former spouses. Until 1984, courts were directed to exercise these powers in such a way as would keep the parties, so far as it was practicable and (having regard to their conduct) just to do so, in the financial position in which they would have been if the marriage had not broken down. Since the removal of that direction by the Matrimonial and Family Proceedings Act 1984 (see further Maclean and Eekelaar, chapter two, this volume), there has been no statutory guidance as to the aims of the process. The legislation simply states that first consideration is to be given to the welfare of any child of the family, that the court is to take all the circumstances of the case into account (referring judges to a list of specific but non-exhaustive factors), and that the court is to consider the desirability of an order that will achieve a clean break between the parties, that is to say, to bring to an end any ongoing financial ties between them. In line with the shift away from a divorce law based on fault to one based on the breakdown of the marriage, the conduct of the parties is rarely a factor that will influence the division of the parties’ assets, save where it is ‘obvious and gross’18 or directly affects the assets available for division.19 14

Child Support Act 1991, and Children Act 1989 s 15 and Sch 1. Contrast the responses given to more specific questions regarding particular legal entitlements: Barlow et al (2001). 16 See Barlow et al (2008). 17 See generally the attitudes of those interviewed by Cooke, Barlow and Callus (2006) about what the law should be, both for spouses and cohabitants; a majority of British Social Attitudes respondents think that cohabitants should pool their resources: Barlow et al (2008) table 2.4; see Harris-Short and Miles (2007) 3.5 for discussion of community of property and joint ownership systems. 18 See Wachtel v Wachtel (No 2) [1973] Fam 72. 19 As in the case of financial or litigation misconduct: see eg M v M (Financial Misconduct: Subpoena Against Third Party) [2006] 2 FLR 1253. 15

Sharing Lives, Dividing Assets 7 The principles underpinning the court’s exercise of discretion were discussed by the House of Lords in the cases of White v White20 and Miller; McFarlane21 and subsequently by the Court of Appeal in Charman v Charman.22 Three bases for reallocation were articulated: need, compensation, and sharing. From a theoretical perspective this proliferation poses a problem, in that each of the bases presupposes a slightly different model of marriage; from both a practical and a theoretical perspective, there is the challenge of deciding whether, and if so when, needs should take precedence over sharing, and how compensation fits into the court’s assessment. There have also been debates as to what exactly is to be shared: does the principle apply only to assets acquired as a result of the parties’ joint efforts during the marriage, or does it extend to other assets acquired before,23 during, or after24 this period?25 At present the position would appear to be that the court will first consider the needs of the parties, and that any assets owned by either party, whenever acquired, may be used to ensure that such needs are met.26 In lower-income cases, meeting the needs of the parent with primary care for the children of the marriage may well result in that party receiving more than half of the assets. If there is a surplus after the needs of the parties are met, then the sharing principle comes into play. This principle does not, however, simply dictate that all assets, or even all assets acquired during the marriage, are to be divided equally between the parties: the duration of the marriage and the contributions of the parties may suggest that a different distribution would be fairer.27 As the Court of Appeal put it in Charman v Charman, ‘property should be shared in equal proportions unless there is good reason to depart from such proportions’.28 This flexible approach reflects the way in which the modern courts have consistently eschewed a formulaic approach.29 They have also been reluctant to develop the idea of compensation introduced in Miller; McFarlane. While there is now a greater awareness of the impact of sacrificing a career,30 such sacrifices tend to be addressed by means other than including a specifically compensatory element

20

[2001] 1 AC 596. [2006] UKHL 24. 22 [2007] EWCA Civ 503. 23 A clear answer in the negative was given in B v B (Ancillary Relief) [2008] EWCA Civ 543; see also McCartney v Mills McCartney [2008] EWHC 401 (Fam). 24 See eg the debates regarding post-separation property in Rossi v Rossi [2006] EWHC 1482 (Fam); S v S [2006] EWHC 2339 (Fam); H v H [2007] EWHC 459 (Fam). 25 See also Vaughan v Vaughan [2007] EWCA Civ 1085, in which Wilson LJ drew a distinction between equality of assets and equality of outcome. 26 See eg S v S (Non-Matrimonial Property: Conduct) [2006] EWHC 2793; S v S (Ancillary Relief: Importance of FDR) [2007] EWHC 1975. 27 See eg S v S (Non-Matrimonial Property: Conduct) [2006] EWHC 2793; NA v MA [2006] EWHC 2900; McCartney v Mills McCartney [2008] EWHC 401 (Fam). 28 [2007] EWCA Civ 503 [65]. 29 See eg H v H [2007] EWHC 459 (Fam). 30 See eg Lauder v Lauder [2007] EWHC 1227 (Fam); VB v JP [2008] EWHC 112 (Fam). 21

8 Joanna Miles and Rebecca Probert in that party’s share of the assets.31 In lower-income cases the ongoing effect of sacrifices made during the marriage (for example giving up paid employment to care for the children of the marriage) will be dealt with as an aspect of that party’s needs;32 in higher-income cases the sharing principle may produce more than the spouse would have earned as a result of paid employment,33 and, or alternatively, the loss of earning capacity will be recognised through a ‘generous assessment’ of that party’s needs.34 In other cases the principle is not seen as relevant, either because the spouse had not developed a career before the marriage,35 or because giving up work was a ‘lifestyle choice’ rather than a necessity.36 And even if the principle is seen as appropriate, courts may struggle with its application to the case before them: one judge suggested that it was neither possible nor desirable to attempt to ascertain the career that the spouse would have had.37 As yet, therefore, there have been few cases in which the concept of compensation has played an independent role in the court’s deliberations.

The Law of Property and Trusts: Dividing the Assets when Cohabitants Separate Decisions made during the lifetime of the relationship have a profound, if not necessarily anticipated, effect on the parties’ property rights on separation. Most couples who have purchased a home together in the past decade will be jointly registered as legal owners of that property and will have made an express declaration as to their beneficial interests in the property38 (even if some of them may not have realised the effect of so doing39). Such a declaration is determinative of their rights—in the absence of fraud or common mistake—regardless of their respective contributions.40 The recent House of Lords decision in Stack v Dowden41 made clear that if a couple purchase a home in joint names without making any such declaration, the starting point will be that they too intended that the beneficial title should be held jointly,42 and strong evidence will be required 31

See eg H v H [2007] EWHC 459 (Fam). See eg Lauder v Lauder [2007] EWHC 1227 (Fam). 33 CR v CR [2007] EWHC 3334 (Fam). 34 See eg Lauder v Lauder [2007] EWHC 1227; VB v JP [2008] EWHC 112. 35 NA v MA [2006] EWHC 2900. 36 S v S (Non-Matrimonial Property: Conduct) [2006] EWHC 2793. 37 P v P [2007] EWHC 779 (Fam). See also VB v JP [2008] EWHC 112 (Fam) and CR v CR [2007] EWHC 3334 (Fam). 38 Since 1998 the relevant Land Registry forms have invited those purchasing a home in joint names to indicate whether they hold the beneficial title as joint tenants, or as tenants in common in equal shares, or on some other trusts. However, the Land Registry will not reject a form that fails to complete this declaration. 39 See Douglas, Pearce and Woodward, ch 7, this volume. 40 Goodman v Gallant [1986] 1 FLR 513. 41 [2007] UKHL 17. 42 See also Abbott v Abbott [2007] UKPC 53. 32

Sharing Lives, Dividing Assets 9 to show that, ‘unusually’, a different arrangement was intended. By contrast, if the home was purchased in the name of one party (or was owned by him or her before the relationship began), then the onus is on the other party to establish a beneficial interest under the law of trusts. This may be done by showing either that the claimant made a financial contribution to the initial purchase price or that there was an agreement, arrangement or understanding between the parties on which the claimant has relied to his or her detriment.43 The former may give rise to a resulting trust (which will yield an interest in the property proportionate to the original contribution), or, more probably in the domestic context,44 will lead the court to infer a constructive trust on the basis that there was a common intention that the beneficial interest should be shared, the financial contribution establishing the necessary element of detrimental reliance. The extent of the parties’ interests under a constructive trust will be determined by reference to the whole course of dealing between the parties. Such distinctions lead to some stark differences in outcome. In Fowler v Barron,45 for example, the property had been conveyed into joint names, but during the 17 years of the parties’ relationship it was Mr Barron who made all the payments under the mortgage and who met other household expenses. Miss Fowler’s salary was used for more transient items, such as holidays, meals out, and clothes in particular for the children and herself. The court’s interpretation of this was that, despite the lack of a joint bank account, the parties had effectively pooled their assets and did not care who paid for what, or how much; the inference was that they had not intended that their different contributions should lead to them having different interests in the property. Each party was thus entitled to 50 per cent of the equity in the property. By contrast, in James v Thomas,46 Miss James moved into a home already owned by Mr Thomas, and he was reluctant to convey it into their joint names. Although she worked unpaid in his business for the better part of their 15-year relationship and the mortgage was paid from the profits of that business, the court held that there was no common intention that she should have an interest in the property, and that her contributions were simply attributable to the fact that she and Mr Thomas ‘were making their life together as man and wife’.47 She was therefore not entitled to any interest in the home.48 Whether these outcomes truly reflected the intentions of the parties is open to question: Mr Thomas acknowledged that it would be fair for Miss James to receive some interest in the property, while Mr Barron argued that, while he had intended that Miss Fowler should inherit the property if he pre-deceased her, he had not intended that she should be entitled to half if they separated. This is not an 43

Lloyds Bank v Rosset [1991] AC 107. At least where the parties are sharing a home: see eg Adekunle and Others v Ritchie [2007] EW Misc 5 (EWCC); cf Laskar v Laskar [2008] EWCA Civ 347. 45 [2008] EWCA Civ 377. 46 [2007] EWCA Civ 1212. 47 ibid [36] (Chadwick LJ). 48 See also Morris v Morris [2008] EWCA Civ 257. 44

10 Joanna Miles and Rebecca Probert unusual view: it has been suggested that many in this situation would be perfectly happy for their partner to inherit all their property in the case of their death, but are far less willing to split their assets equally in case of separation.49 The sharing that is part of a loving relationship is often fiercely resisted by the parties when that relationship has come to an end, whether they were married or cohabiting (as is evident from the research of Lewis et al in this volume). The law, however, takes the view that while the parties’ intentions regarding the shares each has in the property may change over time, those intentions—and so those shares—must at any given point in time be the same for all purposes.50

Financial Provision for Children whose Parents Live Apart In order to complete the picture of the current law relating to the distribution of wealth on relationship breakdown, it is necessary briefly to outline the law governing provision for children. This law applies almost identically to children of married and unmarried parents.51 However, given the very different background of provision/division of property between the adults, described above, the overall outcomes for formerly marital and cohabiting households and the children within them are very different. The Child Support Act 1991, as amended, declares the responsibility of each parent to maintain his children,52 regardless of the nature of the parents’ relationship. In the case of a ‘non-resident parent’ (a parent with whom the child does not principally live), this responsibility is met by the payment of child maintenance calculated by reference to a formula which (subject to possible variation on limited grounds set out in the Act) requires payment of a given percentage of that parent’s income, depending on the number of children to be supported. Recent changes effected by the Child Maintenance and Other Payments Act 2008 mean that child maintenance is now viewed principally as a matter to be resolved by (non-binding) private agreement between the parties.53 The statutory agency, its formula and enforcement powers are intended to provide a fallback where such agreement is unattainable or breaks down. Basic child maintenance (generally made by way of regular payments) is not a matter for the courts, except where agreed by the parties and enshrined in a 49

See Douglas, Pearce and Woodward, ch 7, this volume. Contrast the findings of Douglas, Pearce and Woodward, ch 7, this volume, who find that parties may have precisely the sort of multiple intention that the law does not accommodate. 51 The only key distinction pertains to the responsibility of step-parents: only an individual who is married to (or a civil partner of) a child’s legal parent may be required to maintain the child; someone who has merely lived with that parent may not: see the concepts of ‘parent’ in Children Act 1989 Sch 1 para 16(2); and ‘child of the family’ in Matrimonial Causes Act 1973 s 52(1). 52 Child Support Act 1991 s 1. Only ‘qualifying children’, as defined by ss 3 and 55 of the Act, are owed this duty. 53 Prior to these reforms, if the parent with care was claiming a means-tested benefit, the Child Support Agency’s involvement was effectively mandatory. 50

Sharing Lives, Dividing Assets 11 ‘consent order’.54 However, the courts have exclusive jurisdiction to order various forms of capital provision ‘for the benefit of the child’.55 Crucially, that qualification has been interpreted in such a way as effectively prevents the outright transfer of property to the parent with care. Generally, any capital transferred must either be spent for the child’s benefit during the child’s minority (for example on depreciating assets such as a car for taking the child to and from school and other activities, or ‘white goods’ for the child’s home), or, in the case of the transfer of a house, the property must revert to the other parent once the child attains independence.56 Provision for dependent children of a formerly married or cohabiting couple can therefore offer no long-term or direct security for the parent with care. But, as we have seen, that limitation of these remedies is of far greater significance for former cohabitants (who have no personal financial remedy and may have no claim in property law) than for former spouses (who have the benefit of financial relief under the matrimonial legislation, whatever their property rights may be).

Future Reform? As increasing numbers of couples cohabit outside marriage and as more children are born outside marriage, reform providing statutory financial remedies between cohabitants—already available in a number of other jurisdictions57—has been actively considered. The Law Commission for England and Wales recently published its recommendations for a scheme of remedies between separating cohabitants who had either had a child together or lived together beyond a minimum duration (fixed somewhere between two and five years), unless they had executed a valid ‘opt-out agreement’ excluding the operation of the scheme and (perhaps) made their own arrangements.58 Under this scheme, the courts would have largely the same powers to reallocate assets as they enjoy on divorce, but those powers would be exercisable on a different basis. The Law Commission concluded that, given broadly observable differences between many cohabiting and marital relationships (including the substantial heterogeneity of the cohabiting population; the lack of express, public commitment to the relationship; and various patterns

54 The courts also have exclusive jurisdiction (see Child Support Act 1991 s 8) to order periodical payments for meeting educational expenses, costs associated with extra needs created by a child’s disability, and ‘top-up’ payments where the non-resident parent’s income exceeds the jurisdictional limit of the statutory agency responsible for child maintenance: Children Act 1989 Sch 1; Matrimonial Causes Act 1973 s 23(1)(d) and (e). In the last case, in particular, payment may include a ‘carer’s allowance’ to cover the parent with care’s needs as carer: Re P (A Child: Financial Provision) [2003] EWCA Civ 837. 55 These powers are exercisable by the courts under various statutes. In the case of married parents who are divorcing, jurisdiction would usually be exercised under the Matrimonial Causes Act 1973; in the case of unmarried parents, under the Children Act 1989 Sch 1. 56 A v A [1995] 1 FCR 309. 57 Most recently, Scotland: Family Law (Scotland) Act 2006. 58 Law Commission (2007). See also Law Commission (2006) for its earlier consultation paper.

12 Joanna Miles and Rebecca Probert of money management and property ownership), provision between cohabitants should not be based on the needs of the parties or a norm of equal sharing. Instead, the court would exercise a principled discretion with restitutionary and compensatory purposes designed to respond to the enduring economic impact of the parties’ contributions to their life together. The court would therefore aim broadly to reverse gains (‘retained benefits’) made by one party as a result of the other’s contributions, such as the latter’s direct or indirect financial contribution to the acquisition of property held in the former’s sole name; and to share the burden of losses (‘economic disadvantage’) suffered by one party for the benefit of the family, for example, the loss of future earnings arising from having given up paid employment to care for children.

The Role of Empirical Data in Developing the Law The consequence that, following Stack v Dowden, the proceeds of sale of the family home will be divided 50:50 in most joint ownership cases between cohabitants on separation may seem to resemble the approach in the context of ancillary relief in the wake of cases such as White v White and Miller; McFarlane. There is, however, a difference: the ‘sharing principle’ established in the latter context rests upon the view that the assets accrued during the ‘partnership’ of marriage should as a matter of principle be shared, while the presumption that equity follows the law in the context of the family home rests on the assumption that this is what such couples would usually have intended. The role that individual intentions play in the context of ancillary relief remains a moot point,59 whereas evidence as to the parties’ intentions may play a crucial role under the law of trusts.60 This brings us on to the crucial distinction between empirically based presumptions and normative judgments. The principles developed and asserted in the recent House of Lords decisions may be put into these two inter-related categories: (i) empirically based presumptions about what parties to particular relationships may be taken to intend in particular circumstances: for example, that couples who purchase a home in joint names intend to hold it in equal shares; that couples who manage their money separately may not have that intention. (ii) normative judgements about what consequences ought to be taken to flow from particular situations: for example, that marriage is a partnership 59 For example, pre-nuptial agreements retain a rather limited role, just one factor for the court to take into account in the exercise of its discretion, although in some cases an agreement may be a factor of ‘magnetic importance’: Crossley v Crossley [2007] EWCA Civ 1491. 60 See eg Williamson v Sheikh [2008] EWCA Civ 990 (house in sole name of male partner, intentions gauged from an unexecuted declaration that female partner would be entitled to £10,000 plus 60% of proceeds).

Sharing Lives, Dividing Assets 13 which should entitle the partners to equal shares of certain property, regardless of need or financial contribution; that ‘special contributions’ should nevertheless receive recognition; that needs arising during a marriage (and perhaps to some extent afterwards) should generally be met by the other spouse; that some measure of ‘compensation’ ought to be afforded to recompense one spouse for economic disadvantage incurred, typically as a result of sacrificing paid employment to care for home and family. To some extent, these empirically based presumptions and normative judgements articulated by the House of Lords were consciously based not just on the courts’ assessment of the position of individual litigants who have appeared before them or on ‘common sense’ views adopted by the courts over time, but also on observations yielded from social science data, including socio-legal research. The speeches of Baroness Hale, in particular, are distinctive for and often enriched by her use of such material.61 Whether legal development is to be driven by reference to how people actually think or behave in their relationships, or by reference to what society considers the law ought to be seeking to achieve, or (perhaps most likely) a combination of both, social science data and analysis have central roles to play in informing the deliberations of appellate courts, law reformers and policy-makers. Such information is always essential to obtaining a proper understanding of the context in which the current law is operating and how that law is impacting on those to whom it applies. For example, data regarding participation in paid employment during relationships and about the economic disadvantage experienced following relationship breakdown by those who have been absent from the labour market may readily prompt the preliminary conclusion that remedies should be provided in response to this hardship.62 Determining precisely how that should be achieved, on what basis and in which cases may, of course, be more difficult to resolve, but research data may have a valuable contribution to make here too. In a rather different vein, data about how couples deal with the financial implications of separation, and about the obstacles that they encounter in the process, may prompt us to re-examine both procedural and other contextual aspects of dispute resolution and the extent to which complexity in the substantive law may itself impede settlement. Moreover, without a clear understanding of the social context in which the law is to operate, the law of unintended consequences is likely to rule the day.

61 In Miller, McFarlane and Stack v Dowden she refers to Pahl (1989), Ermisch (2000), Barlow et al (2001), Lewis (2001), Arthur et al (2002) and Douglas, Pearce and Woodward (2007) (cf the contribution by the latter authors to this volume (ch 7)). See also her extra-judicial writing: eg Hale (2004) draws on Rake (2000). 62 Although see remarks in the conclusion to this chapter regarding the relative responsibility of private family law and other areas of law and policy to deal with the issues explored in this volume.

14 Joanna Miles and Rebecca Probert The importance of empirical data can only grow as those considering legal reform in this field explore schemes based not on the exercise of a wholly individualised, strong discretion, with which many are becoming disenchanted,63 but on a more ‘standardised’ basis, such as community of property regimes, guidelines for wealth distribution,64 or ‘structured discretion’ of the sort recently recommended for cohabitants by the Law Commission.65 If a (more or less) standardised scheme is to ‘fit’ the population to which it is to apply—to supply a satisfactory measure of ‘average rather than individualised justice’66— its development must be fully informed by data on a wider range of issues such as: patterns of property ownership, money management, employment and so on, and what (if anything) these patterns imply about parties’ intentions, commitment and so on; the degree of disparity in earning capacity or standard of living between partners following separation; and public attitudes towards the provision of remedies in particular cases and about the appropriate basis for remedies.67 It is also important that the judges in our appellate courts should refer to social science data when developing the current law, as they have done in recent cases such as Stack v Dowden.68 All too often their Lordships are not referred to key findings bearing on the empirical questions which they inevitably have to address in the course of their decision-making.69 One major obstacle is, of course, the accessibility of such data: those presenting the case to the court will not usually be au fait with literature in non-legal disciplines, and even research assistants employed by the court may find it difficult to discover and interpret papers written by economists, sociologists and psychologists. It is equally unlikely, of course, that empirical researchers will be aware of ongoing legal proceedings to which their findings may be highly relevant, and terminological differences sometimes pose a barrier to mutual understanding. There is thus a role for intermediaries to make empirical research from one discipline more accessible—both physically and intellectually—to another. It must, however, be conceded that no distillation of the existing research evidence will ever be value-free. Indeed, it is clear from academic criticisms of Stack v 63 See eg the disadvantages of the current Matrimonial Causes Act scheme identified by the Law Society (2003) [71]; concerns reported in the wake of Miller, McFarlane by the Court of Appeal in Charman v Charman (No 4) [2007] EWCA Civ 503 [120]; and proposals for presumptive formulae for the calculation of periodical payments awards: eg Eekelaar (2006). 64 See for example the Canadian spousal support guidelines: Department of Justice Canada (2008). 65 Law Commission (2006, 2007). 66 Rogerson and Thompson (2005) 10. 67 Cf the view of Maclean and Eekelaar, in ch 2, this volume at p 34 regarding the difficulty today of identifying shared expectations and values on which reform might be based. 68 See for example Harris-Short and Miles (2008), Probert (2008a, 2008b), and Douglas, Pearce and Woodward, ch 7, this volume. 69 For example, there is no reference in Stack v Dowden to research into couples’ money management practices, data which have been central to many commentators’ criticisms of the case—see previous note.

Sharing Lives, Dividing Assets 15 Dowden that the research data cited in that case is open to interpretations different from those adopted by the House of Lords. Since such different interpretations can generate different conclusions about the appropriate response from the law, there is a need for constant evaluation of the interplay between such research evidence and judicial decision-making—an example of which appears in this volume, in chapter seven, by Douglas, Pearce and Woodward. THE ISSUES AND THE DATA

As Maclean and Eekelaar indicate in chapter two of this volume, much research to date has focused on divorced or separating couples. But every separating couple was once an intact family, and the choices made while the relationship was subsisting will obviously be a significant constraint on the options available on relationship breakdown. Part II of the book, ‘Work, Money and Property within Intimate Relationships: Expectations and Actions’, examines data relevant to these questions. Problems originating during the relationship are particularly acute where one or both of the parties are in debt. There has to date been relatively little discussion of this aspect of relationships or its legal implications: for example, we now recognise a loose concept of ‘matrimonial property’—should we also recognise more explicitly the reality of ‘matrimonial debt’, and if so how should we distinguish it from personal debt?70 Finney’s discussion of debt within relationships and the links between debt and relationship breakdown in chapter six highlights key background data about both indebtedness and saving habits, and their association with particular types of relationship and life events such as relationship formation, the arrival of children, and relationship breakdown. In addition, the House of Lords in both Miller; McFarlane and Stack v Dowden placed some importance on the way in which the couple dealt with their assets during the relationship, both when dividing assets on divorce and when discerning the parties’ intentions regarding their property. But without empirical data it is difficult to know what forms of behaviour are unusual, or what light such behaviour can throw on the parties’ intentions. The assumption that Mr Stack and Ms Dowden were unusual in keeping their finances separate is clearly contradicted by several studies. The chapters by Vogler (chapter four), Burgoyne and Sonnenberg (chapter five), and Douglas, Pearce and Woodward (chapter seven) cast valuable light on this area, showing in particular that what we might instinctively assume certain money management practices indicate about parties’ intentions regarding ownership and use of their money might be incorrect. Of course, an individual’s behaviour may reflect constraints as well as their preferences, as is evident from the analysis by Scott and Dex in chapter three 70 cf the findings of Cooke et al (2006), where respondents to their qualitative survey were largely not supportive of debts, as well as assets, being treated as joint as between spouses.

16 Joanna Miles and Rebecca Probert of the division of labour in the context of the UK labour market.71 Given the difficulties of changing the underlying conditions that produce such inequality, it is likely that it will persist for the foreseeable future, and it is therefore necessary to consider its effect both during relationships and upon separation. Unequal earnings may, for example, affect the way in which couples manage their money and generate differing concepts of ‘equality’ in this sphere, as chapter four by Vogler, and chapter five by Burgoyne and Sonnenberg show. For some, inequality in income is seen as mandating inequality in contributions, with each contributing according to their ability;72 for others, equal contributions are made despite differences in earnings. But as the next section shows, contributions, in whatever proportions, do not always translate into a share of the assets, equal or otherwise, upon relationship breakdown. In the third part of the book, ‘Dividing the Assets on Relationship Breakdown’, the discussion of married and cohabiting couples of necessity diverges, owing to the very different legal regimes that apply to each on divorce and separation. Douglas, Pearce and Woodward in chapter seven draw on empirical research to highlight how difficult it is to ascertain the ‘common intention’ required by the law of trusts, particularly by reference to the factors identified by the House of Lords in Stack v Dowden. They show that the law has set itself a daunting task in seeking to give effect to parties’ intentions, given the difficulties encountered in practice in attempting to ascertain from the parties’ conduct of their relationship any shared intention specifically regarding property ownership. Lewis, Tennant and Taylor in chapter eight identify those disadvantaged by the current law relating to cohabitants and consider whether the Law Commission’s recommendations for reform in this area would do any better at addressing those individuals’ problems. Their study shows, in particular, how legal ownership (specifically) rather than the needs of the parties (or beneficial entitlements under the law of trusts) currently dominates the way in which assets are divided when a cohabiting relationship breaks down. By contrast, upon divorce needs assume far greater importance: indeed, as Hitchings shows in chapter nine, the needs of the parties, and the range of practical options available to meet those needs, are of crucial importance in determining the nature of any settlement in most everyday cases. The emphasis on needs may help to achieve a degree of consistency in the disposal of these cases, despite the lack of guidance in the legislation. Further research is necessary in this area, not least given the concerns expressed by the Law Society in 2003 about uncertainty of outcome, which may vary between different county courts.73 ‘Local knowledge’ may be extremely important in practice, which poses a particular problem for the growing number of litigants in person appearing

71

And see further La Valle, Clery and Huerta (2008). See also Stack v Dowden [2007] UKHL 17 [91], where Baroness Hale refers to the idea of ‘each doing what they could’, as a form of organisation that could then reflect an intention that the property be equally shared. 73 Law Society (2003) [60]–[70]. See also Hitchings, ch 9, this volume. 72

Sharing Lives, Dividing Assets 17 before the family courts, who will almost certainly lack the insider knowledge garnered by repeat players. Moreover, public misconceptions about the role of the law in this context may hinder a speedy settlement, and Dowding’s chapter ten sets the issue of settlement in its practical context, identifying the many opportunities which parties have— and are strongly encouraged by their legal advisors to use—to reach agreement without recourse to litigation, and the cost implications of their choices in this respect. As the chapters in this volume by Vogler (chapter four), Burgoyne and Sonnenberg (chapter five), and Douglas, Pearce and Woodward (chapter seven) indicate, couples often find it difficult to discuss money and property during their relationships—or easy to avoid having to do so. Inevitably if their relationship breaks down, such discussions become unavoidable and the earlier the couple can be helped to resolve their disagreements, the lower will be the costs of their dispute, both financial and emotional. But in the case of cohabitants (where there are no judicial powers of the sort enjoyed on divorce to adjust property rights and where the special matrimonial procedures described by Dowding in chapter ten do not apply), by the time the couple do belatedly address the issue of who owns what (and so who leaves the relationship with what), it may be too late for one party to guard against disadvantage of the sort discussed by Lewis, Tennant and Taylor in chapter eight. One feature, however, is common to both cohabiting and married relationships, especially where there are dependent children, and that is inequality on relationship breakdown arising in particular from decisions made during the relationship about each party’s participation in paid employment. The impact of these decisions on the income of men and women after divorce and the end of cohabitation is made manifest by Fisher and Low in chapter eleven, based on an analysis of several sweeps of the British Household Panel Survey. Their conclusion that income recovers over time gives some cause for optimism, although it is clear from their analysis that in many cases income recovery following divorce is achieved by repartnering rather than by improvements in personal income from employment. While this may guard against poverty following divorce, repartnering may be said simply to substitute a new dependency for the old one, and as such fails to address what for many is the underlying problem. Moreover, as Fisher and Low demonstrate, repartnering is less likely for older women and for women with children, leaving them economically disadvantaged unless able substantially to increase their labour market participation, which few appear to do. Price’s chapter on pensions (chapter twelve) builds on this theme of postdivorce inequality, providing powerful evidence that under current conditions divorced women do not and cannot catch up in pension building, and that, contrary to government’s expectations, pension sharing on divorce is rare rather than routine, leaving divorced women facing poverty in old age. The reasons for the rarity of pension-sharing orders are hard to diagnose, although some anecdotal evidence suggests that pensions are not routinely included in parties’ schedule of assets (‘Form E’) on divorce, so potentially rendering that valuable

18 Joanna Miles and Rebecca Probert asset ‘invisible’ and not a ‘matrimonial asset’ to be shared. The assumption may be that women will create their own pension fund by ‘downsizing’ once the children reach independence, or that they will protect themselves via repartnering; and we know from research that some women are principally motivated to ensure their short-term security (for example in securing accommodation for themselves and the children) to the neglect of their long-term interests.74 But as Fisher and Low’s data show, the latter assumption may be misplaced, and the inherent uncertainty about prospects of repartnering and about the security of any new relationship makes it difficult to factor that possibility into any individual settlement. Moreover, the principles developed by the House of Lords in Miller; McFarlane have clear application to pension savings: savings accumulated during the marriage should be treated as matrimonial property and in principle be brought into account in dividing those assets pursuant to the ‘sharing’ principle; and women’s disadvantage in relation to pensions offers a clear example of the sort of economic disadvantage to which the ‘compensation’ strand is intended to alert us. However, we also lack information about the characteristics of the tiny minority of cases in which such orders are made. This is clearly an area ripe for further investigation and, given the uncertainty that surrounds future state provision beyond retirement age, is a matter of considerable importance as the proportion of older women who are divorced, and living longer, increases. Fisher and Low’s data also provide material against which they evaluate alternative principles on which financial remedies following relationship breakdown might be based—comparing income-based measures (reflecting the impact of the relationship on each party’s human capital) with consumption-based measures (based on needs or expected living standard)—and so for analysing the outcomes in and the incentive effects created by Miller; McFarlane. This theme—how to choose the most appropriate basis for financial relief—is developed further in the contrasting contributions of Dnes and Barlow in the final part of the book, ‘A Rational Approach?’. Barlow in chapter fourteen draws on recent social attitude surveys and demographic changes to support her argument in favour of a ‘bottom-up’ approach to law reform: a reconfiguration of remedies on relationship breakdown according to whether the parties (married or not) had children, with a focus on needs-based remedies between couples with children and on greater autonomy for childless couples. Meanwhile, Dnes’ chapter thirteen offers an economist’s view of the law governing financial and property disputes on divorce and the end of cohabitation, demonstrating the potential for economic analysis to provide greater clarity to discussions in this area, for example by articulating welfare goals more explicitly and highlighting the different incentive effects of various settlement rules, including the introduction of financial remedies at the end of cohabitation.75

74 75

Douglas and Perry (2001). See generally Dnes and Rowthorn (2002).

Sharing Lives, Dividing Assets 19 These chapters also provide competing views on the ability of the law to influence behaviour within the family context, and how different research methodologies can be used to detect such influences at work. This discussion has been described by Barlow and others as being concerned with ‘legal rationality’, that is the idea that people respond to the framework of family regulation created by the law and order their intimate relationships in line with the law’s expectations.76 Barlow argues, in light of data yielded from various research projects conducted by herself and other socio-legal scholars, that legislators in the family-law sphere are in danger of making a ‘rationality mistake’, overestimating law’s ability to influence behaviour in the family field. She considers the evidence of people’s changing attitudes to relationships and property and how well or otherwise this fits with the law’s expectations, arguing that the legal and financial consequences of marriage rarely figure in the decision whether to marry (at least for first marriages). Even assuming that the parties have an accurate understanding of the law, to marry for such pragmatic reasons is seen by many as distasteful, despite the potentially far-reaching economic consequences of this decision when the relationship ends, whether by separation or by death. Dnes, by contrast, explains the insights that can be provided by positivistic, quantitative analysis of longitudinal samples (such as that undertaken by Fisher and Low), arguing that how people explain their own behaviour to qualitative researchers may not necessarily correspond with independent indicators about what people are in fact doing, and highlighting a number of studies identifying significant changes in certain types of behaviour, including partnering behaviour, in response to legal change. This debate about the influence of law on family behaviour is undoubtedly one that will continue and requires further examination. It may be that some laws, such as those relating to welfare benefits and tax credits77 have greater influence over family practices than those relating to financial settlements following relationship breakdown,78 not least given what to the individuals concerned is the relative complexity and perceived remoteness of those laws from decisions about marriage and conduct within ongoing relationships. WHERE SHOULD WE GO FROM HERE?

One clear finding that emerges from the empirical research is the great diversity of practices within families. Whether such diversity militates against devising proposals for reform on the basis of shared values, as Maclean and Eekelaar

76 In the language of economists, ‘rationality’ is simply concerned with observations about the consistency or otherwise of a subject’s decisions, and is entirely unrelated to the desirability of those decisions, the information on which they are based, or their compatibility with or conformity with a particular policy goal or objective perception of the subject’s best interests, given the current legal framework. 77 See eg Anderberg (2008). 78 Although see Mechoulan (2006), who found that the rules governing allocation of property on divorce had a stronger effect on the divorce rate than the grounds for divorce.

20 Joanna Miles and Rebecca Probert suggest in chapter two, is a moot point. The surveys discussed by Barlow in chapter fourteen suggest that there is still a considerable degree of common ground in the wider population, and a solution that does not resonate with individuals’ values may be difficult to implement in practice. As noted above, there are difficulties in ascertaining the parties’ intentions regarding their property from their money management practices during their relationship. There is also a more fundamental question to be posed, which is how far the law should be guided by the parties’ intentions when dividing assets at the end of a relationship. To date, as we have seen, the answer to that question has differed according to the nature of the relationship, rather than the extent of the disadvantage experienced by the individual in question. This has resulted in a situation in which very similar contributions may result in very different outcomes (as demonstrated by Lewis, Tennant and Taylor, in chapter eight, who compare the results that their cohabitants might have expected had they been married with the actual arrangements that were made). Yet although individual experiences differ, some common trends can be observed. While we have, as a society, departed from the pure version of Lloyd Cohen’s profile of marriage,79 it remains the case that women tend to earn less than men and spend more time on unpaid household tasks. Whether the first causes the second, or vice versa, is difficult to determine; still more difficult is the task of devising a solution to encourage greater equality in both spheres.80 The positive story of women’s enhanced earning capacity and labour market participation has a negative consequence: the extent of downward mobility for those who move out of the labour market even for a short period is greater for today’s women than it was for previous generations. And although remaining out of the labour market to care for children and other dependants should be respected as a choice, this does not mean that the financial detriment of doing so should fall on the carer alone. A division of responsibilities that works within the intact family may well leave the partner who has not pursued a career vulnerable on relationship breakdown (as demonstrated by Fisher and Low and by Price in chapters eleven and twelve). To the extent that the other party—whether married or cohabiting—has benefited from the contribution of the one who provided care or other non-tangible contributions, then there is a justification for the court to intervene as between the parties themselves. But on what basis? Underlying much of the discussion in this volume is the concept of ‘equality’. But this is a multi-faceted concept which could support a variety of solutions to the problems experienced on relationship breakdown. Should equal sharing of assets arise only following equality of contribution? Is contribution to be regarded as equal only where it is of equal value, 79

Cohen (1987). See eg recent comments of Nicola Brewer, chief executive of the Equalities and Human Rights Commission, regarding the negative impact that improved maternity rights have had on women’s progression in the labour market: Bennett and Ahmed (2008). 80

Sharing Lives, Dividing Assets 21 or made as equal proportions of individual wealth? Are domestic contributions assumed to carry equal value? Or ought we instead to focus on substantive equality of outcome, and if so how is that to be achieved? This brings us back to the ideas of equality of consumption level (needs) or ‘compensation’.81 Studies suggest that the division of responsibilities during the relationship has a direct impact upon both the reduced earnings of married women82 and the enhanced earnings of married men.83 The courts’ lack of enthusiasm for the concept of ‘compensation’ is to a large extent attributable to their perceived inability to put a value on the loss that needs to be compensated. Human capital projections of the sort necessary to put a precise figure on future earnings losses and reduced pensions are unfamiliar territory for family lawyers, and potentially costly tools for parties in dispute to have to provide.84 However, that there will have been some disadvantage to the primary carer of children will in many cases be selfevident, and the research by Fisher and Low reveals what the average effects of divorce on the parties’ income are, at least for the general population. One possible way forward might be to use that data as a starting point for deciding compensation awards on divorce, rather than, as at present, starting from an assumption that there is no loss to be compensated, and requiring individualised proof by the applicant. Although a ‘one-size-fits-all’ approach is likely to be inappropriate in this complex area, there is scope for a more standardised approach to ancillary relief than currently exists.85 While the focus of this volume is on informing the development of private law entitlements and obligations in light of the research data, it is important to retain a wider perspective and an awareness of the limited horizons and capacities of private family law. Many of the issues arising from the work discussed here go to fundamental issues of social rather than merely inter-personal justice, in particular relating to equality of opportunity. Where individuals’ or families’ choices about their behaviour—in particular regarding participation in paid employment, pensions and other saving, and indebtedness—are constrained by social and labour market conditions, created, fostered or maintained by a particular web (or absence) of legal regulation and other policy levers, we must critically examine those areas of law and other tools of governance. Moreover, to the extent that disadvantage and hardship experienced on separation can be characterised as a product of 81 See Miles (2008) for a recent discussion of how a principle of compensation should be used as a tool for achieving equality of outcome on divorce. 82 See eg Bryan and Sevilla Sanz (2008), who suggest that the timing and nature of housework has an impact on women’s wages, since they tend to be responsible for daily tasks that have to be fitted around the working day. 83 Korenman and Neumark (1991), based on US evidence; although compare Bardasi and Taylor (2005) in the UK context, who use a different methodology and find a rather smaller premium: see further Fisher and Low, ch 11, this volume. 84 But note, for example, the evidence provided in a recent Scottish cohabitation case, CM v STS [2008] CSOH 125 from [284]. 85 Compare, for example, the work of Rogerson and Thompson (2005) and Department of Justice Canada (2008).

22 Joanna Miles and Rebecca Probert social constraints rather than private preferences—and of socially beneficial work in caring for dependants—we need to ask whether responsibility for supporting the carer should fall on the other partner at all. It could instead be argued that it is a matter (to that extent) of public rather than private responsibility.86 Reform beyond the traditionally-understood ambit of ‘family law’ (in particular, in the fields of labour law and social security law) may well—in the long term— provide a more satisfactory remedy to many of the problems to which family law is currently endeavouring, not without difficulty in some instances, to provide an adequate response.87 This is not to say, however, that family lawyers and law-makers can or should simply rely on other areas of the law to produce greater equality between men and women and so mitigate the impact of relationship breakdown. There will always be many who choose to prioritise caring for the family above paid employment, and family law must respond to the economic consequences of relationship breakdown in a way which reflects as closely as possible the real-life experiences of those involved. For this ongoing project to succeed, we need to ensure a constant cycle of empirical research to feed into the ‘life-cycle’ of legal development. First, research has a vital role to play in informing the creation or development of the applicable legal regime. This should include research on the experience of jurisdictions whose legal regimes we are contemplating borrowing: without such research, there is a real danger of ‘faulty transplant’ or of assuming on the basis of the face of the legislation and case law that all is well in practice. Then, once reform is enacted, research is needed to monitor how the new law is operating in practice, in part to identify any unintended consequences, not least where people behave in ways that do not conform with the law-makers’ expectations. Meanwhile, it remains necessary to maintain a programme of research into contemporary attitudes, norms, and behaviours in the fields of money management, labour market and household work participation, and so on; and, as discussed in the previous paragraph, to revisit the appropriate division of responsibilities and functions between family law and other areas of law and policy with the power to affect behaviour in these spheres. And so, as norms and behaviour change, such research will have a role to play in shaping the next reform of family law. BIBLIOGRAPHY Anderberg, D (2008) ‘Tax credits, income support and partnership decisions’ 15 International Tax and Public Finance 499. Arthur, S, Lewis, J, Maclean, M, Finch, S and Fitzgerald, R (2002) Settling Up: Making financial arrangements after divorce or separation (London, NatCen).

86 87

See eg Fineman (2004). See Price, ch 12, and Scott and Dex, ch 3, this volume.

Sharing Lives, Dividing Assets 23 Bardasi, E and Taylor, M (2005) ‘Marriage and Wages: A Test of the Specialisation Hypothesis’, Working Papers of the Institute for Social and Economic Research, Paper 2005-1 (Colchester, University of Essex). Barlow, A, Burgoyne, C, Clery, L and Smithson, J (2008) ‘Cohabitation and the Law: myths, money and the media’ in A Park et al (eds), British Social Attitudes: The 24th Report (London, Sage). Barlow, A, Duncan, S, James, G and Park, A, (2001) ‘Just a piece of paper? Marriage and cohabitation in Britain’ in A Park et al (eds), British Social Attitudes: The 18th Report— Public policy and social ties (London, Sage). Bennett, R and Ahmed, M (2008) ‘Equality laws “are now holding women back”’, The Times July 14. Bottomley, A and Wong, S (2006) ‘Shared Households: A New Paradigm for Thinking about the Reform of Domestic Property Relations’, in A Diduck and F Kaganas (eds), Feminist Perspectives on Family Law (Abingdon, Routledge-Cavendish). Bryan, M and Sevilla Sanz, A (2008) ‘Does Housework Lower Wages and Why? Evidence for Britain’, Working Papers of the Institute for Social and Economic Research, Paper 2008-3 (Colchester, University of Essex). Burgoyne, C, Reibstein, J, Edmunds, A and Dolman, V (2007) ‘Money management systems in early marriage: factors influencing change and stability’ 28 Journal of Economic Psychology 214. Burns, M, Burgoyne, C and Clarke, V (2008) ‘Financial affairs? Money management in same-sex relationships’ 37 Journal of Socio-Economics 481. Clarke, V, Burgoyne, C and Burns, M (2007) ‘Romance, rights, recognition, responsibilities and radicalism: Same-sex couples’ views on civil partnership and marriage’ in V Clarke and E Peel (eds), Out in Psychology: Lesbian, Gay, Bisexual, Trans and Queer Perspectives (Chichester, Wiley). Cohen, L (1987) ‘Marriage, Divorce and Quasi Rents; or, “I Gave Him the Best Years of My Life”’, 16 Journal of Legal Studies 267. Cooke, E, Barlow, A and Callus, T (2006) Community of Property: A Regime for England and Wales (Bristol, The Policy Press). Department of Justice Canada (2008) Spousal Support Advisory Guidelines, http://canada. justice.gc.ca/eng/pi/pad-rpad/res/spag/ssag_eng.pdf. Dnes, A and Rowthorn, R (eds) (2002) The Law and Economics of Marriage and Divorce (Cambridge, CUP). Douglas, G, Pearce, J and Woodward, H (2007) ‘Dealing with Property Issues on Cohabitation Breakdown’ 37 Family Law 36. Douglas, G and Perry, A (2001) ‘How parents cope financially on separation and divorce—implications for the future of ancillary relief ’ 13 Child and Family Law Quarterly 67. Eekelaar, J (2006) ‘Property and Financial Settlements on Divorce: Sharing and Compensating’ 36 Family Law 754. Ermisch, J (2000) ‘Personal Relationships and Marriage Expectations’, Working Papers of the Institute for Social and Economic Research, Paper 2000-27 (Colchester, University of Essex). Fineman, M (2004) The Autonomy Myth: A Theory of Dependency (New York, The New Press). Hale, B (2004) ‘Unmarried Couples in Family Law’ 34 Family Law 419.

24 Joanna Miles and Rebecca Probert Harris-Short, S and Miles, J (2007) Family Law: Text, Cases and Materials (Oxford, OUP). —— (2008) Family Law: Text, Cases and Materials website, www.oup.com/uk/orc/ bin/9780199277162/resources/updates/ch03_0409.pdf. Korenman, S and Neumark, D (1991) ‘Does Marriage Really Make Men More Productive?’ 26 Journal of Human Resources 282. La Valle, I, Clery, E and Huerta, MC (2008) Maternity Rights and Mothers’ Employment Decisions, Research Report No 496 (London, Department for Work and Pensions). Law Commission (2006) Cohabitation: The Financial Consequences of Relationship Breakdown, Law Com CP 179 (London, TSO). —— (2007) Cohabitation: The Financial Consequences of Relationship Breakdown, Law Com No 307 (London, TSO). —— (2008) Tenth Programme of Law Reform, Law Com No 311 (London, TSO). Law Society (2003) Financial Provision on Divorce: Clarity and Fairness (London, Law Society). Lewis, J (2001) ‘Debates and Issues regarding Marriage and Cohabitation in the British and American Literature’ 15 International Journal of Law, Policy and the Family 159. Masson, J, Bailey-Harris, R and Probert, R (2008) Cretney’s Principles of Family Law, 8th edn (London, Sweet and Maxwell). Mechoulan, S (2006) ‘Divorce laws and the structure of the American family’ 35 Journal of Legal Studies 143. Miles, J (2008) ‘Charman v Charman: making sense of need, compensation and equal sharing after Miller/McFarlane’ 20 Child and Family Law Quarterly 378. Pahl, J (1989) Money and Marriage (Basingstoke, Macmillan). Peel, E and Harding, R (eds) (2008) Regulating Sexuality: Contemporary Perspectives on Lesbian and Gay Relationship Recognition, special issue of Sexualities. Probert, R (2008a) ‘Equality in the family home?’ 15 Feminist Legal Studies 341. —— (2008b) ‘Cohabitants and joint ownership: the implications of Stack v Dowden’ 38 Family Law 924. Rake, K (2000) Women’s Incomes over the Lifetime (London, Cabinet Office). Rogerson, C and Thompson, R (2005) Spousal Support Advisory Guidelines: A Draft Proposal (Canada, Department of Justice). Smart, C, Mason, J and Shipman, B (2006) Gay and Lesbian ‘Marriage’: An Exploration of the Meanings and Significance of Legitimating Same-sex Relationships (Manchester, Morgan Centre for the Study of Relationships and Personal Life), www.socialsciences. manchester.ac.uk/morgancentre/research/gay-lesbian-marriage/. Vogler, C (2005) ‘Cohabiting couples: rethinking money in the household at the beginning of the twenty-first century’ 53 Sociological Review 1. Vogler, C, Brockmann, M and Wiggins, R (2006) ‘Intimate relationships and changing patterns of money management at the beginning of the twenty-first century’ 57 The British Journal of Sociology 455. —— (2008) ‘Managing money in new heterosexual forms of intimate relationship’ 37 Journal of Socio-economics 552. Weeks, J, Heaphy, B and Donovan, C (2001) Same Sex Intimacies: Families of Choice and Other Life-experiments (London, Routledge). Williams, F (2004) Rethinking Families (London, Calouste Gulbenkian Foundation).

2 The Perils of Reforming Family Law and the Increasing Need for Empirical Research, 1980–2008 MAVIS MACLEAN AND JOHN EEKELAAR

INTRODUCTION

T

he impetus for bringing together in this volume these papers on recent developments in the making of financial arrangements on relationship breakdown lies in the recent decisions of the House of Lords, that is, in the impact of judicial decision-making. This chapter seeks to fill out the picture of the development of family law by looking at the role of legislation over the last 30 years, and highlights the growing need for an evidence base for law reform. The family law policy process in England and Wales is driven by the government of the day: if the administration has a comfortable majority in the House of Commons and is ‘hungry’ to legislate after a long period in opposition, as was the case with the New Labour government of 1997, then a flurry of legislative activity is likely to ensue. This was indeed the case, even though the current Brown administration is perhaps interested in consolidation rather than forging ahead. Nevertheless at the time of writing we are still in a period of ‘courageous’ and proactive law-making, and family policy has developed with a strong focus on the needs of children (see, for example, DCSF, 2005). By contrast, reforming the law relating to adult family members has proved more difficult, even where the existing law has an adverse impact on children.1 When concern arose about the financial disparity between the outcomes of parental separation for children of cohabiting parents compared with married parents, the Law Commission carefully and thoroughly addressed the question of the need for further financial regulation for separating cohabitants (Law Commission, 2007) and advised legislation. But the government announced in March 2008 that it does not propose to take action for the time being. Making family law, to quote Ira Ellman (Ellman, 2003), is hard … in this case too hard. And it is getting 1

See Hale (2004).

26 Mavis Maclean and John Eekelaar harder. As society becomes more fluid and more diverse, the chances of creating new family law which will not offend the private values and expectations of a significant number of potential voters become slimmer and slimmer. The struggles over the Family Law Act 1996 are still remembered. This legislation began as a sensible attempt to tidy up our rather confusing grounds for no-fault divorce (breakdown of the marriage needing often to be evidenced by clearly blameworthy behaviour including adultery2) and promote less adversarial ways of resolving related disputes, but became a battleground for different interest groups. The government of the day was not strong in the House of Commons and so accepted amendments from widely disparate sources, until the resulting Act resembled Joseph’s Coat of Many Colours. As is well known, its main provisions have never been implemented and it awaits partial repeal. If making family law is so hard, will any government ever be brave enough to try? If we want to encourage ministers to move forward and make sure that our family law is fit for purpose as family structures change and diversify, how can we encourage them to have courage? Alternatively, if we fear that making new family law is so difficult that a proactive government may make things worse, how do we alert them to the possible elephant traps which stand in their way? The answer, whichever position we start from, must be empirical research. This chapter will first describe how research has contributed to the development of the law which regulates money and property arrangements on divorce or separation, and then set out some suggestions for the way ahead, looking not only at the socio-economic position of men and women undergoing this process but also at the ways in which they reach their decisions about financial matters. PRAGMATISM AND THE RISE OF EMPIRICISM

Until the Divorce Reform Act of 1969 and related reform of financial provision on divorce in the Matrimonial Proceedings and Property Act 1970 came into effect in 1971 (these provisions now being found in the Matrimonial Causes Act 1973), the powers of the courts in relation to finance and property on divorce and the basic principles under which they worked had not been changed significantly since the nineteenth century. The husband had a duty to provide income support, but the court had no power to make any order concerning his property, although it could do so against the property of a guilty wife. A major step forward in 1971 led (see Eekelaar, 2000) to gender-neutral provision, and the articulation of a goal: that of placing the parties as far as it was practicable to do so in the position they would have been in had the marriage not broken down (taking into account conduct). In order to achieve this aim, the courts acquired a general power to transfer property from one spouse to the other. At this time, the purpose of the law was to discourage 2 Contrast the no-fault separation ‘facts’ on which irretrievable breakdown of marriage, necessary to obtain a divorce, can be proved: Matrimonial Causes Act 1973 s 1(2)(d) and (e).

The Perils of Reforming Family Law 27 divorce, by holding financial responsibilities to continue in the event of the ending of the marriage, and by penalising conduct (although the courts swiftly eroded this element limiting the definition to obvious and gross and misconduct).3 While this kind of normative purpose lay behind the divorce reforms the policy-maker’s task was clear: to ensure that the law would support the generally agreed aim of protecting marriage. But as divorce law became altogether more pragmatic, and the goal on separation became making the best of the resources available to the parties bearing in mind the needs of the children, then it became more important to develop a better understanding of how family finances were operating. Two social changes in particular affected the way in which the law operated, and became a matter to be studied and understood. First, the increasing rate of home ownership made the power to transfer the home more significant. If the husband’s interest in the matrimonial home could be transferred to the wife either outright or for the period while the children were dependent, the wife’s bargaining position was strengthened immeasurably. Secondly, women’s increasing participation in the workforce, although their work was often part time with low pay, provided a different perspective on the need to remain in an unhappy marriage from that of the woman with no independent resources at all. During the 1970s, public debate on the financial consequences of divorce developed, as calls for reform were prompted by the ‘no-fault’ approach, combined with the possibility that ‘innocent’ men might be losing the homes that were in their names and on which they were paying the mortgage. The Campaign for Justice in Divorce (CJD) was established in 1978, and ran a skilful media campaign on behalf of divorcing men and their second wives. The Law Commission again addressed the issue, and in 1980 published its discussion paper (Law Commission, 1980) agreeing that the current objective of the courts of trying to avoid loss for all parties was simply not working. But to avoid any extension of the judicial discretion which could lead to variation in decision-making and uncertainty (see Barrington Baker and Eekelaar, 1977) the factors to be taken into account in making financial arrangements on divorce, while giving priority to the interests of the children, also included the suggestion that the desirability of the parties becoming self-sufficient should be formulated as a positive principle: that is women should work, and a clean financial break between spouses should be the aim. The situation was a striking example of the failure of the government to respond to deep-rooted changes to family patterns, particularly the growth in the number of separated and reconstituted families and the increasing complexity of cross-household parental responsibilities, leading to the emergence of an angry interest group. The CJD made effective use of anecdotal evidence, but there was little scientifically based information to go on. Were men being hard done by? Were wives rejecting their husbands, stealing their houses and children, and going

3

See Wachtel v Wachtel [1973] Fam 72.

28 Mavis Maclean and John Eekelaar on to live what was referred to as ‘the life of Riley’ at the expense of the happiness of their former husbands and their new families? As social science developed, and a better understanding of the state of society began to be part of general knowledge, it is not surprising that the matter of rising divorce rates and post-divorce finance began to attract the attention of researchers. In 1980, John Eekelaar and I began our study of the financial consequences of divorce. When we began to speak with other ‘divorce’ researchers in the UK we were isolated, as the main thrust of such research was focused on the psychological and social consequences of divorce. But we were encouraged to find researchers in other jurisdictions tackling the same issues, (see Weitzman and Maclean, 1992). In 1981, with the support of the Economic and Social Research Council, we carried out a survey of a nationally representative sample of divorced men and women and found clear evidence, as Lenore Weitzman had found in the United States (Weitzman, 1985), of the economic detriment which women experienced after divorce. As soon as our findings began to emerge (Eekelaar and Maclean, 1986) we were targeted by the CJD, who circulated a pamphlet to all Members of Parliament entitled ‘Mud on her Gown … a critique of Mavis Maclean’. But our findings were, we hoped, the basis for a better-informed approach to the matter, as it was clear that the major factor in the economic difficulties experienced by the women was their responsibility for children (see also Funder, 1992). The typical household situation pre-divorce was two adults with one-and-a-half jobs and dependent children. After divorce, there were two households: one household with one earner and possibly a second partner earning something also; and one household, possibly in the former matrimonial home, with one half-time earner and dependent children. Women’s wages were low, child care expensive and hard to find, and women therefore needed either to earn enough to cover the costs of child care and other work-related expenses such as travel, or to stay at home and live on state benefits plus whatever periodical payments were made by the former husband. These might be minimal if he had left the wife and children in the former matrimonial home (Eekelaar and Maclean, 1986). With hindsight these findings seem totally predictable. But at the time they were highly original, based on sound empirical research, and offered a counter-balance to the anecdotal claims of the pressure group of the day, CJD. We were asked to give evidence to a special standing committee of the House of Commons during the passage of the 1984 legislation. And the Act which followed (the Matrimonial and Family Proceedings Act 1984) was more restrained in approach than early drafts had indicated, as the minimal loss principle was removed, but the courts were to give first consideration to the needs of the children. The so-called ‘clean break’ was not to be a positive principle. Instead, the court should merely consider whether self-sufficiency was appropriate. Despite the isolation of the researchers in the UK, we felt supported by the comparable findings emerging from other jurisdictions. And when the government undertook a larger study to replicate our work, our findings were consistently confirmed (see Weitzman and Maclean, 1992, following a conference in

The Perils of Reforming Family Law 29 1989). The 1984 Act—which amended the Matrimonial Causes Act 1973, still in force today—gained public acceptance, and remains the basis for our regulation of financial arrangements after divorce. We suggest that the firm research base made a major contribution to the good fit between the aims of the legislation and the contemporary situation of divorcing and repartnering families. The more pragmatic the intentions of the drafters become, the more important it becomes to have a good understanding of the social context into which the legislation is inserted.

THE CHANGING POLICY FOCUS: FROM ADULTS TO CHILDREN

Although the divorce arrangements described above concerning spousal support and property division remain in place, they were rapidly overtaken in public debate by the issue of child support. In this matter, legislation in the form of the Child Support Act 1991 was made rapidly and without research support, at the insistence of the then Prime Minister, Margaret Thatcher. The results may be considered as unsuccessful in that the law was universally unpopular, both with men who were asked for larger sums of money and with women on welfare whose households, without any disregard for maintenance they received, failed to secure any benefit. Child support assessments were subject to long delays and punitive emergency assessment, and compliance was low. The legislation was swiftly amended and is currently being replaced by the Child Maintenance and Other Payments Act 2008, setting up a system which should benefit from a better evidence base following a large research programme carried out by the Department for Work and Pensions (Wikeley et al, 2008). The government now has a much clearer picture of the characteristics of the families involved in cross-household parenting, and in particular the rapidly changing job, housing, and household circumstances experienced. The problems associated with the Child Support Act 1991 are not, of course, caused by of lack of research, but rather by the intransigent nature of the problem of persuading non-resident parents to part with much larger sums of money than they had been used to. But had there been a better background knowledge base there would at least have been warning of some of the issues which were going arise, and at least some aspects of the 1991 legislation might have been handled differently. For example, requiring detailed information as to non-resident parents’ actual housing costs might have been modified if it had been appreciated how often these were likely to alter in the period after separation, and how informal they tended to be and difficult to evidence. But, in this case, the empirical work which preceded the reform concentrated on poverty among lone parents. There was little attempt to examine the way the judicial part of the process was working, nor enough attention paid to investigating the circumstances of the non-resident parents , a notoriously difficult group to reach for research purposes. And the work on understanding the underlying social norms about parental obligations came too late to inform this first attempt at solving

30 Mavis Maclean and John Eekelaar the problem through statute (see Maclean and Eekelaar, 1997, an empirical study of the attitudes towards and support in place between parents who had divorced, separated, or never lived together), although it has been used in the policy work in preparation for the 2008 Act. This study clarified the difference between men’s and women’s attitudes to social and biological parenting. For the parents with care, mainly women, support for the first family was a clear priority, and any second family was expected to take second place in the call on the limited resources of the non-resident parent. But the non-resident parents, mainly men, were torn between meeting the needs of their first and second families and prioritised their current household, which might include step-children, that is children of their new partner, as well as new children, that is children of the new relationship. The difficulties which followed the Child Support Act 1991 made it clear that, although quantitative research into the circumstances of lone parents had been rigorous and helpful, there was also a need for qualitative work to provide information about the attitudes and expectations of all the parents concerned. Serial partnering with children is an expensive exercise, and there is a continuing division of opinion as to whether the cost should be born by the individuals concerned (in which case licence to repartner might be made contingent on earning capacity) or whether the state should step in. The only broad consensus is around the need to protect the children from the poverty associated with living in a single-parent household.

MOVING FROM DIFFERENTIATION TOWARDS INCLUSIVENESS IN ALLOCATING RESOURCES ON SEPARATION

We noted earlier how the early divorce legislation differentiated between men and women, and how the Matrimonial Proceedings and Property Act 1970 (now Part II of the Matrimonial Causes Act 1973) for the first time removed this distinction in allocating resources after the marriage ended. Before the child support legislation, there had always been a differentiation between the approach to children whose parents had married and children whose parents had lived together, or never formed a common household. This distinction in providing financial support was swept away by the Child Support Act 1991, which dealt with the financial obligations of all non-resident parents to all of their children. But the distinction remains between the approach to the division of property and other assets to be made at the end of a marriage compared with those made at the end of cohabitation, and this necessarily impacts indirectly on the children of such relationships. This differentiation between the protection offered by the law to the children of separating cohabitants and the children of divorcing parents, which flows from the lack of remedies for the benefit of parents with care outside the marital context, and the lack of public awareness of the distinction between the legal position of wives and that of cohabiting women, were key stimuli to calls for legislation

The Perils of Reforming Family Law 31 to extend the protection offered to divorcing mothers and children to separating mothers and children. Brenda Hale’s article in Family Law puts the case forcefully (Hale, 2004). When the Law Commission received the reference to examine the issue, they approached the issue with rigour and energy and began by collating the available research evidence on the characteristics of these families (age, housing status, children, work status, earning capacity and so on). The importance of this background information to any proposed change in the law is now no longer in question. But the thrust of the story of the increasingly close relationship between research and policy-making is consistent. The development of the research knowledge base available to policy-makers, despite concerns about lack of funding and lack of research capacity (see Genn et al, 2006) is at last bringing together not only the necessary large-scale survey work but also the detailed and nuanced qualitative research which will enable us to see how individuals manage family change, particularly when uncertainty or dispute arises. To summarise, increasingly diverse and flexible family forms mean that there are no longer clear, universally held assumptions to be made about family circumstances; the increasing pragmatism of family law reform, aiming to offer management of family matters rather than abstract justice based on moral or religious principles, means that it becomes ever more important for the policy-maker to understand what individuals expect and value.

THE WAY AHEAD

We have now reached the position where active discussion and the development of proposals to regulate the financial outcomes of separation, as opposed to divorce, have added to the growing interest in looking again at the checklist of factors in section 25 of the Matrimonial Causes Act 1973, as amended in 1984, which guide the court’s discretion in granting financial remedies. It might be argued that as the differences between marrying and cohabiting couples seem to be diminishing (the cohabitants are becoming older, economically better off, experiencing longer partnerships, and approaching the level of stability of the married population (see Law Commission, 2006: Part 2)) there is little point in attempting to deal with one group without the other. The other stimulus for looking again at the situation of the married group lies in the decisions made recently in some rather unusual ‘big money’ cases where the abundance of assets to be disposed of seems to have encouraged an abundance of ideas about how this might be done. In the past, by far the largest group of divorces were those affecting couples with limited means, where the courts used their discretion to attend first to the need to house the children and the parent with care (nine times out of ten the mother). The Child Support Act 1991 then required the non-resident parent to make an administratively calculated regular payment to the parent with care for the children. In most cases, these two aspects of the arrangements exhausted the available resources. There is still some spousal support of limited duration to support the

32 Mavis Maclean and John Eekelaar mother of young children or to enable the former wife to find work and adjust to the new situation. But long-term spousal support is becoming more and more rare. There may be adjustments to be made with respect to pensions, but often the amounts available do not justify the cost of the actuarial work required to make the allocation. A roof and child maintenance is the general expectation. But recent cases have highlighted the confusion around what kind of outcomes the courts are seeking. The legal arguments range around the importance of the duration of the marriage which has ended, the extent of the responsibility to provide for a former spouse’s needs, how to share the combined benefits accruing from the marriage and, conversely, how far one party should be compensated by the other for any detriment incurred during the marriage. In the unusual case of White v White,4 where there were no dependent children and the considerable assets were agricultural, Lord Nicholls said: [W]here the assets exceed the financial needs of both parties why should the surplus belong solely to the husband? … If, in their different spheres each contributed equally to the family, then in principle it matters not which of them earned the money and built up the assets … Before reaching a firm conclusion and making an order along these lines, a judge would be well advised to check his tentative views against the yardstick of equality of division. As a general guide, equality should only be departed from if, and to the extent that, there is good reason for doing so.5

This case and the following cases of Miller; McFarlane6 are discussed elsewhere in this volume. But the issue for those seeking to respond to current concerns about the lack of certainty facing those who divorce, and the extent of judicial discretion, is how to respond to the disproportionate impact which these cases have had on the average high-street case.7 Given that any case reaching the House of Lords is unlikely to be a typical run-of-the-mill case, but instead one with unusual circumstances and sufficient assets to be worth the argument, the decision is unlikely to help the average case. Such a decision may even cause additional problems by appearing to set out a clear principle, for example, equality of division, where there is an excess above what is required to meet needs. Fisher’s (2002) small study of high-street solicitors in the year following White v White8 did indeed find that clients of modest means were expecting all assets to be sold and shared. In low- to average-income households this would mean that the family home must be sold and that, while the male earner would be able to start again in the mortgage market, a woman with child-care responsibilities and no qualifications would find it all but impossible to get started as an owneroccupier. With the decline in public housing provision and the lack of low-cost privately rented accommodation, this is a serious issue. Under the existing rules, 4 5 6 7 8

[2001] AC 596. ibid at 605–8. [2006] UKHL 24. Although compare Hitchings, ch 9, this volume. [2001] AC 596.

The Perils of Reforming Family Law 33 the court has a duty to give first consideration to the needs of the children, which in practice requires attention to be paid to keeping a roof over the heads of the children and the parent with care and ensuring that there are sufficient funds to maintain the position. The judiciary struggle for clarity and consistency in these wealthy cases. The high-street solicitor tries to create a survival package for his client, stretching oneand-a-half incomes to support two households. Is it possible to set out meaningful rules to help people struggling to reorder their lives in such a wide variety of circumstances? Or has the judicial discretion which we have accepted for so long given us the flexibility to keep a roof over the children’s heads and enable the mother to gather herself together and adjust to life after marriage, without killing the goose who lays the golden egg, that is the major earner? Members of the Bar have recently called for more certainty, arguing that they cannot advise their clients in negotiation when the parameters are so broad. But whenever a scheme is suggested as a basis for decision-making, there is little sign of agreement. At a seminar held in Oxford in March 2007, John Eekelaar put forward a scheme based on duration of marriage to a group similar to the contributors to this volume and including members of the senior judiciary and legal practitioners. While many supported the idea of a framework, none would accept any of the proposals put forward. What is the policy-maker to do, given professional concerns but relatively little information other than anecdote as to what kinds of values and expectations men and women have about the financial arrangements to be made on divorce, or about what happens in practice? Nor should we ignore the continuing debate about the values and expectations at the end of cohabitation, and the relatively limited information about what happens to both opposite-sex couples and samesex couples, who may or may not have entered into a civil partnership. We suggest that the first step would usefully be a substantial quantitative research project to examine the socio-economic characteristics of the adult separating population, both those who had married or entered into a civil partnership and those who had been cohabiting (and those referred to by demographers as ‘Living Apart Together’: Haskey, 2005). We need to understand the history of such relationships, the patterns of money and asset management and ownership, what the the parties’ financial expectations would have been if the relationship had not broken down, and the obligations of the parties to others including children and other dependants of this and previous relationships. We need to understand what they thought should be the financial outcome of their separation, what this expectation was based on, and how close it is to what the current family justice system would accept. Finally, we would need to discover what arrangements were made, how they were made, what professional intervention there was, and whether the arrangements have been upheld and whether they are thought to be satisfactory. Such a survey would have to begin from a general population sample, and could not be completed in less than two years, but it would provide the baseline data for future policy-making and evaluation. In the absence of shared values and

34 Mavis Maclean and John Eekelaar expectations it is not possible to develop new law that will be widely acceptable or enforceable. In the absence of the information set out above we do not even know how the current system is working and whether it is already working as well as can be hoped, given that the divorcing/separating population is unhappy and unlikely to be enthusiastic about any legal framework. The child support legislation is an example of an attempt to legislate for the benefit of children, an uncontentious goal. The support obligation is almost universally accepted. But very few nonresident parents find their own assessment satisfactory. For the policy-maker, family disputes are dangerous territory. But while we wait for this research to be funded and completed, we suggest that it may also be helpful to look in depth at the way in which separating couples approach the making of financial arrangements. Some manage to make their own arrangements informally, others seek the advice of family solicitors to reach agreement, and a minority fail to resolve matters without the help of the courts and the family Bar. These conflicted cases may tell us a great deal about the kinds of expectations men and women bring to this situation, what kinds of issues lead to dispute, and how the justice system tries to contain and manage these conflicts and help the parties to reach a practicable arrangement which takes into account their needs, their resources, and above all their children.

RESOLVING FINANCIAL DISPUTES WITH THE HELP OF THE FAMILY BAR

At the Oxford Centre, we have carried out research (Eekelaar, Maclean and Beinart, 2000) into the divorce work of family courts and lawyers that reveals something of how individuals in dispute approach the issues and how they are guided towards settlement by legal advisers, only rarely requiring adjudication by the court. We studied family solicitors, and observed the careful supportive process by which these lawyers lead their clients towards settlement. The lawyer negotiates with his client to reach a position which falls within the parameters which he knows from experience to be acceptable to a court. If the other party’s solicitor follows a similar process, then both parties should be within reach of agreement by the time their representatives begin to negotiate with each other. Difficulties arise when a client will not accept the lawyer’s advice and presses for a position which is beyond the scope of what a judge could accept. Our current qualitative research (Maclean and Eekelaar, 2009) looks at the work of the family Bar and includes observational data on financial disputes at the end of a marriage. These detailed accounts are based on our observation of the day-to-day work of counsel in ancillary relief cases, and provide a starting point for considering potential reform of section 25 of the 1973 Act. Would rules help? We observed big-money cases, middle-income cases, and cases where there were no resources to argue about but rather the apportionment of debt and access to welfare benefits. In the big-money category, one case we observed turned on the belief of the wife that her husband had additional substantial assets which he

The Perils of Reforming Family Law 35 had not disclosed. A great deal of time and money was spent in trying to pursue this issue, until she eventually accepted that more efforts could be made but that they would be unlikely to add to the existing information, and her legal advisers finally convinced her that it was probable that all substantial assets had now been discovered. How the assets were to be divided was not the cause of the conflict, but what the assets comprised. The skills required from the barrister were meticulous attention to the detail in the multiplicity of financial papers, perhaps more akin to the forensic work of an accountant rather than a lawyer. In a second big-money case, the issue again was not the proportion of the assets to go to each party, which was not at issue. The dispute centred rather on the way the assets were to be valued, as they included hedge funds which could not be given a precise valuation at the time of the divorce hearing. One side wanted to agree the date at which a valuation should be made for the assets to be divided in a certain ratio. The other side wanted a fixed amount regardless of the precise date of valuation. Of the cases where the assets were modest and the debts considerable, the matters in dispute again included questions about disclosure of information. One wife could simply not believe that all the money raised by a second mortgage had been spent, but according to the evidence provided it clearly had. In addition, there were questions about how to make best use of the assets available to the parties. The dispute centred around whether the husband would be able to keep his job and hold on to the former matrimonial home, which would then increase in value to the benefit of both parties in the longer term, or whether it should be sold and the proceeds shared now even though this would be of little benefit to either side as the wife was in receipt of benefits and her entitlement would be threatened. In a second low-money case, the problem was lack of financial knowledge and experience. Both parties had modest incomes (he earned £20,000 per annum and she earned £3,000 per annum) but each had run up debts totalling £50,000 on a number of different credit cards. He had moved in with a new partner. She hoped to hold on to the former matrimonial home, while he wished to sell and settle his debts. The judge asked how she might finance her plan, and she said that friends would help but she had no definite figures or firm commitments. The court required the sale and equal sharing of the equity in the house. It is hard to see how any more precise rules about property division could have helped in any of these cases, all of which were based in the facts of the matter and not in the application of legal doctrine. In the big-money cases, resources were more than adequate to meet needs, and this is where questions about the role of compensation or equality may emerge, although we did not see this. In the ‘no-money’ cases, resources cannot even meet needs, never mind do any more than this, and so we saw questions about disclosure and how to best manage debt and welfare benefit entitlements. Perhaps the middle-money cases, where both parties are trying to continue to support their present standard of living but with divided resources, is where the division of assets might benefit from further regulation or guidance on principle? But even here the major task is to meet the needs of both parties for housing and

36 Mavis Maclean and John Eekelaar the income necessary to maintain it. Here, the arguments are about house prices in various areas, about the distance from the children’s schools, and, increasingly often, about the need for two homes large enough for the children to spend time comfortably in as part of shared parenting arrangements. When contact visits did not involve staying more than the odd night in school holidays, makeshift arrangements were more acceptable. But in a middle-class, middle-income family with shared care, there is now a need to fund a bedroom in each home for each child, an additional burden on the financial arrangements. These brief accounts may serve to illustrate the problems that would arise on trying to devise principles for property division on divorce or separation. It would be possible to abandon the aim of meeting need and move towards a formulaic division of the assets. But, as can be seen from these examples, the conflicts which reach the courts arise from lack of trust in the other party, and involve refusal to accept the other side’s account of the financial position without full forensic enquiry, supported by the authority of the court, to obtain disclosure of bank and credit card details and the questioning of statements from accountants and employers about actual and potential earnings and expenses. The task of the legal advisers we observed was not to argue points of law but to obtain the trust of their client and then to painstakingly comb through the figures and reach an agreed statement of the assets on the table, that is, the size of the ‘pot’. The division of the pot was usually a less contentious matter. Any attempt to establish clearer guidance about the factors to be taken into account in assessing need would have to consider for example whether this should include provision for shared parenting. Any attempt to regulate how to share resources available after needs have been met would have to consider whether the ratio should be 50:50, or a sliding scale depending on contributions or length of relationship or numbers of children or a shared understanding of the wife’s contribution to family finances. Any codification of how to look at the situation where needs have been met and compensation made for any financial loss resulting from the relationship becomes the issue, must consider, for example, whether the loss was due to any other cause such as illness, although this might be included if caused by the spouse (for instance as a result of passive smoking), but excluded if genetic and unknown at marriage. Taking into account the length of the relationship would not help with establishing the facts or increasing trust between the parties. As Ira Ellman has said, ‘in view of the inherently uncertain nature of the outcomes of legal interventions in family law, it matters less what rules are chosen than that whatever is done is clear and applied with consistency, for that at least will be fair’ (Ellman, 2003). To conclude, pointing out that any attempt to legislate for the regulation of the financial arrangements to be made at the end of a relationship is difficult is hardly necessary. What we have tried to show in this chapter is how the task has changed over the last 25 years, and how these changes have made a firm knowledge base for the policy-maker increasingly necessary. In a homogeneous, stable society with agreed norms and values, law can be made according to agreed principle. For

The Perils of Reforming Family Law 37 many years judges followed the practice of the ecclesiastical courts in awarding the petitioning wife roughly one-third of her husband’s income (Barton, 1957). Now that family forms are complex, changing, flexible, and not encompassed within a single value system, law-making has become more pragmatic and concerned to manage disputes so that people can get on with their lives rather than the more abstract pursuit of truth and justice. If we are to be pragmatic and flexible, we need to know who we are dealing with, what they want, and how to best meet their often conflicting needs.

BIBLIOGRAPHY Barrington Baker, B, Eekalaar, J, Clarke, K and Raikes, S (1977) The Matrimonial Jurisdiction of Registrars (Oxford, Centre for Socio-Legal Studies). Barton, JL (1957) ‘The Enforcement of Financial Provisions’ in RH Graveson and FR Crane (eds), A Century of Family Law (London, Sweet & Maxwell). Department of Education and Skills (2003) Every Child Matters, Cm 5860 (London, HMSO). Eekelaar J (2000) ‘The End of an Era?’ in S Katz, J Eekelaar and M Maclean (eds), Cross Currents: Family Law and Policy in the US and England (Oxford, OUP). Eekelaar, J and Maclean, M (1986) Maintenance after Divorce (Oxford, OUP). Eekelaar, J, Maclean, M and Beinart, S (2000) Family Lawyers, The Divorce Work of Solicitors (Oxford, Hart Publishing). Ellman, I (2003) ‘Why making family law is hard’ 35 Arizona State Law Journal 699. Fisher, L (2002) ‘The unexpected impact of White—taking equality too far?’ 32 Family Law 208. Funder, K (1992) ‘A Proposal for Reform’ in L Weitzman and M Maclean (eds), Economic Consequences of Divorce (Oxford, OUP). Genn, H, Partington, M and Wheeler, S (2006) Nuffield Foundation Report: Law in the Real World: Improving our Understanding of How Law Works (London, Nuffield Foundation). Hale, B (2004) ‘Unmarried couples in family law’ 34 Family Law 419. Haskey, J (2005) ‘Living Arrangements in Contemporary Britain: Having a partner who lives elsewhere and Living Apart Together’ 122 Population Trends 35. Law Commission (1980) The Financial Consequences of Divorce, Cmnd 8041 (London, TSO). —— (2006) Cohabitation: The Financial Consequences of Relationship Breakdown, Law Com CP 179 (London, TSO). —— (2007) Cohabitation: The Financial Consequences of Relationship Breakdown, Law Com No 307 (London, TSO). Maclean M and Eekelaar J (1997) The Parental Obligation (Oxford, Hart Publishing). —— (2009) Family Law Advocacy: How Barristers help the Victims of Family Failure (Oxford, Hart Publishing). Weitzman, L (1985) The Divorce Revolution (New York, Free Press/Macmillan). Weitzman, L and Maclean, M (eds) (1992) Economic Consequences of Divorce (Oxford, OUP). Wikeley, N, Ireland, E, Bryson, C and Smith, R (2008) Relationship Separation and Child Support Study, DWP Research Report 503 (London, Department for Work and Pensions).

3 Paid and Unpaid Work Can Policy Improve Gender Inequalities? JACQUELINE SCOTT AND SHIRLEY DEX

INTRODUCTION

W

omen’s position in the labour market has improved enormously in recent decades, both in terms of employment and in relative earnings. It was predicted that increasing gender equality within the labour market would lead to greater egalitarianism in unpaid work within the household. However, this has proved to be only partially true. Although employment reduces women’s unpaid work, changes in men’s contribution to domestic labour have been slight and uneven. The last decade or so has seen a Europe-wide explosion of interest in the way social policies, labour markets, and motherhood are inter-related (Del Boca and Wetzels, 2007). In Europe, population decline is an issue and forecasts suggest that below-population replacement levels of total fertility will continue, with an average European total fertility rate of about 1.5 through to 2020 (Kohler, Billari and Ortega, 2006). With people living longer, there is concern that the proportion of people who are of working age compared with those who are too young or too old to work (the dependency ratio) will steadily decline. Thus there is an interest in improving the dependency ratio by alleviating family–work conflicts, so that women employees will continue to have children (the labour market of the future), and so that mothers will continue to participate in the labour force. In this chapter we review some of the ways policies under different welfare regimes seek to influence work–family balance. Policy rhetoric in the UK tends to emphasise the importance of individual choice in making decisions about work–family balance, but this masks the degree to which people’s choices are differentially constrained by gender. In particular, unless the unequal division between men and women of unpaid work can be addressed, policies that seek to reconcile paid work and family life are likely to be extremely limited in their capacity to improve gender inequalities both in the labour force and in the realm of unpaid domestic work and care.

42

Jacqueline Scott and Shirley Dex

In order to understand what policies might facilitate or mitigate against change in men’s and women’s division of labour, we review the changes that have occurred in division of labour within couples in Britain over the past decade and explore why gender inequalities in unpaid work are so slow to change. While much attention has been paid to child-care provision and parental leave policies, far less attention has been given to whether policies can be devised that will equalise men’s and women’s contributions to unpaid work (Dex, forthcoming). Our review leads to the rather pessimistic conclusion that, on the basis of the existing evidence, it is unlikely that state policies will have more than a minimal influence in reducing gender inequalities in a couple’s share of unpaid domestic work. However, before examining in more detail the division of labour within couples, and associated policy initiatives, we first consider how far gender equality has been achieved in the UK labour market.

GENDER EQUALITY AND THE UK LABOUR MARKET

There are two very different stories that can be told about labour market changes over the past half-century or so, in terms of gender equality. One story emphasises the positive, while the other offers a somewhat less rosy picture. The positive perspective has plenty of evidence on which to draw: the proportion of women in the labour market has grown markedly; the pay gap has narrowed; notions that a woman’s place is in the home have eroded markedly; and women have overtaken men in numbers pursuing higher education. Half a century ago, the situation was very different. The 1949 Royal Commission on Population report was concerned that the then existing employment bars against married women working were harmful all round—to women, the family, and the community. True, the Commission was hardly giving a ringing endorsement to employed mothers. For example it observes that ‘there is often a real conflict between motherhood and a whole-time career’ (Royal Commission on Population, 1949: 160). Nevertheless it went on to acknowledge that, at least in part, the conflict is due to artificial barriers that restrict the contribution that women can make to the cultural and economic life of the nation. The report urged that a ‘deliberate effort should be made to devise adjustments that would render it easier for women to combine motherhood and the care of a home with outside activities’ (at 160). In 1951, less than a quarter of married women in Britain were in the workforce; by 1991 this was the case for half of all married women, and the proportion continued to rise. By 2001, 65 per cent of married women were in the labour force and there is no longer any difference between the participation rates of married and single women (Gallie, 2000; Scott, 2008). In 1951, 30 per cent of women aged 20–59 were in full-time employment (Joshi, 1989). Forty years later in 1991 this figure had risen only slightly to 34 per cent. In the same period, the extent of women being employed part-time has quadrupled from 11 per cent of all female

Paid and Unpaid Work 43 employees in 1951 to 45 per cent in 1991 (Gallie, 2000). In 2002, 70 per cent of working-age women were economically active, with 42 per cent of those in employment working part time (WEU, 2004). On the positive side, the UK has already surpassed the Lisbon target for a female labour force participation rate of 60 per cent by 2010. But there is little to be complacent about. Women and mothers in particular are often caught in parttime jobs that frequently bring disadvantages in pay and promotion trajectories. Mothers, according to national surveys, overwhelmingly prefer to work part time. However, whether their preferences reflect the fact that shorter hours help women to juggle family and work roles, or whether part-time work is the only realistic option because of lack of child-care alternatives and a traditional gender-role division of labour in the home, is less clear. What is clear from analysis of British Cohort Survey data is that although overall there has been a decrease in the downward mobility of women following childbirth, if women have longer breaks out of the workforce or return after childbirth to a part-time job, the occupational penalties in terms of downward mobility have increased over time. Even professional women experience increasing occupational costs for taking longer periods away from work following the birth of their first child. As Figure 3.1 shows, a teacher born between 1922 and 1936 and between 1943 and 1953 (data from WES—the Women and Employment Survey) had a one in five chance of moving down the occupational scale after taking one year off work; for a woman taking five years off work, this increased to just over a one in four chance. For a teacher born in 1958 (data from the NCDS—National Child Development Survey), there was a one in four chance of moving down the occupational scale following a one-year break, which increased to a one in three chance if the gap was five years.

Predicted probabilities

0.4 0.35 0.3

Year 1

0.25

Year 2

0.2

Year 3

0.15

Year 4

0.1

Year 5

0.05 0 WES

NCDS

Source: Figure from Dex, Ward and Joshi (2008). Figure 3.1: Predicted Probabilities of Downward Occupational Mobility after Childbirth (WES and NCDS teachers by years before first return to work)

44

Jacqueline Scott and Shirley Dex

There has been an overall trend of ever-faster rates of mothers’ return to work after childbirth, as Figure 3.2 shows. For example, 50 per cent of mothers born in 1946 had returned to work by the time their first child was six years old. For mothers born in 1958, 50 per cent had returned within two years after the birth, and of those born in 1970 half had returned after just one year. However, these figures vary greatly for recent cohorts depending on the level of educational qualifications the mothers held, and it is those with higher education who were born in 1958 and 1970 who have returned to work fastest (less than a year after the birth of their first child), whereas those with lower levels of education or no qualifications return at a much slower rate. The analysis of women’s changing patterns of employment over time makes it clear that there has been some improvement in women’s employment prospects. Moreover, these improvements go hand in hand with improvements in women’s tertiary educational qualifications and the lessening of the pay gap. The percentage of women who had tertiary qualifications increased from 11 per cent of those born in 1946, to 25 per cent of those born in 1958, and 32 per cent of those born in 1970; whereas the equivalent percentages for men were 22 per cent, 28 per cent,

16 14

14 13

12

13 12

10 Years

10 9 8

8

6

6

5 4

4

2

2

0

0 1958

1910

1922

1934

1946

1 0 1970

Mother's year of birth All No qualifications Higher qualifications

Source: Figure from Dex, Ward and Joshi (2008). Figure 3.2: Median Number of Years between First Birth and Return to Next Job by Mothers’ Date of Birth and Qualifications

Paid and Unpaid Work 45 and 31 per cent. If we look at the ratio of women’s to men’s hourly wage at age 26 for those working full time, then for those born in 1946 it was 0.63, for those born in 1958 it had increased to 0.84, and for those born in 1970 it was 0.91 (Joshi and Paci, 1998). This is a genuine lessening of the gender pay gap. But there is also a story that is far less rosy. If we look at the average annual gross earnings of graduates, where one might expect to find younger generations of women with the opportunities and inclinations to achieve financial rewards comparable to their male peers, we find this is not the case. Using data from a longitudinal study of over 3,000 graduates who gained their first degrees in 1995, Purcell and Elias (2008) found that young women, even at this early stage in their careers, do not appear to have achieved equal earnings with their male peers and, moreover, that the gender pay gap continues to increase as their careers develop. As we can see in Figure 3.3, women graduates reported full-time gross earnings in their first job after graduation that were on average, 11 per cent less than those of male graduates. Three and half years later the gap had risen to 15 per cent, and by 2002/03 to 19 per cent. So how can this 19 per cent pay gap be accounted for? The top bar on the chart shown in Figure 3.4 gives the unadjusted difference in the annual earnings of male and female graduates in full-time employment seven years after graduation. Each bar beneath this shows the effect on the gender difference in pay of introducing statistical controls for the different relevant factors. Thus controlling for weekly hours alone reduces the differential to about 16 per cent. Next, controls are added for sectors of employment. Women are much more likely than men to work in the public sector, which pays less than the private sector but may have more ‘family-friendly’ working conditions. The sector controls reduce the wage

£35,000 Average annual earnings

£30,000 £25,000

Gender gap = 19% Men

Gender gap = 15%

Women

£20,000 £15,000

Gender gap = 11%

£10,000 £5,000 £0 First main job after 1995 degree

Earnings in 1997/98

Earnings in 2002/03

Source: Figure from Purcell and Elias (2008). Figure 3.3: Average Annual Gross Earnings of 1995 Graduates by Gender

46

Jacqueline Scott and Shirley Dex

Unadjusted gender difference

Adjusted for weekly hours

Adjusted for weekly hours and sector Adjusted for weekly hours, sectors and workplace segmentation Residual gender difference 0

2

4 6 8 10 12 14 16 18 % difference between female annual gross earnings and male annual gross earnings

20

Source: Figure from Purcell and Elias (2008). Figure 3.4: The Combined Effects of Various Factors on the Gender Difference in Annual Earnings of 1995 Graduates Seven Years after Graduation

gap to nearly 11 per cent. Finally the additional impact of gender segregation at the workplace brings the gender gap down to only 7 per cent. There may be some unmeasured workplace discrimination that helps explain a small fraction of the remaining gender inequality in pay. But, beyond the operation of any outstanding workplace inequalities in pay and promotion prospects, a more important explanation lies in the day-to-day practices of unequal division of paid work and caring activities in the household (Gershuny, 2004). The persistence of the pay gap between men and women is likely to be attributable in part to a gender division of labour among couples in paid and unpaid work. HOW FAR HAS THE DIVISION OF LABOUR WITHIN COUPLES IN BRITAIN CHANGED?

The main question addressed in this section is to what extent the divide of paid and unpaid work in couple households in Britain has changed over the last decade. The data used are from the British Household Panel Survey, which is a longitudinal survey of over 5,000 households in Britain. Here we report analysis by Harkness (2008) which concentrates on couples where the women are in their prime working age between 25 and 49. Couples need not be legally married but they must be cohabiting. Thus ‘lat’ couples (those living apart together) are excluded, not only because they are problematic to identify but also because,

Paid and Unpaid Work 47 given the sampling rules of the survey, data for both partners would be available only if they had children. The analysis looks at aggregate patterns of change over time—and does not follow individuals longitudinally. In Figure 3.5 we can see the employment rate for women in couples by whether or not they have children. While employment rates for childless women have always been high, there has been a marked rise in the employment of mothers from 1992 onwards, with the greatest rise among mothers of young children. There is no evidence of any decline in the employment of the male partners in these couple households. The division of paid work in households is very dependent on the presence of young children. While there has been a rise of 6 per cent overall in the proportion of female workers, as Table 3.1 shows, this increase rises to 10 per cent for households with small children (under the age of five). These figures combine dual-earner households (both parties working full time) and the one-and-a-half earner model (where the female partner works part time). This increase in labour may in part reflect the impact of a series of government initiatives supporting maternal employment, including increased support for child care and improved maternity leave. In terms of hours spent on paid work and housework (including cooking, cleaning and doing the laundry, but excluding child care) we can see in Table 3.2 that the paid work hours of women rose on average by two hours per week from 1993/94 to 2003/04; whereas, for women with pre-school children, the increase over the same period was three hours per week to 18 hours in 2003/04. As paid

1 0.9

Employment Rate (%)

0.8 0.7 0.6 0.5 0.4 Couple Kids Couple Kids under 5 Couple no kids

0.3 0.2 0.1 0

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Source: Figure from Harkness (2008). Figure 3.5: Employment Rates for Women in Couples (aged 25–49) (British Household Panel Survey)

48

Jacqueline Scott and Shirley Dex

Table 3.1: Household Work Patterns: Women aged 20–49 in Couples Dual career (both full time)

Modified (male full time and female part time)

Male breadwinner

Female breadwinner

No earner

Sample size

All 1993/94 2003/04 Change

0.38 0.42 0.04

0.27 0.29 0.02

0.21 0.2 –0.01

0.05 0.04 0.01

0.07 0.04 –0.03

3,182 2,760

With children under 5 1993/4 2003/4 Change

0.17 0.2 0.03

0.29 0.36 0.07

0.41 0.38 –0.03

0.03 0.02 –0.01

0.1 0.03 –0.07

871 772

Source: Table from Harkness (2008).

Table 3.2: Hours of Paid and Unpaid Work within Households Paid Work Hours

Unpaid Work Hours

female

male

total

female

male

total

All couple households 1993/94 2003/04 Change

23.4 25.2 1.8

41.4 41.7 0.3

64.8 66.9 2.1

20.8 16.2 –4.6

5.7 5.5 –0.2

26.5 21.7 –4.8

Children under 5 1993/94 2003/04 Change

14.9 17.8 2.9

41.8 42.9 1.1

56.7 60.7 4

24.7 18.7 –6

5.9 5.2 –0.7

30.6 23.9 –6.7

Source: Table from Harkness (2008).

employment rose there was a corresponding decrease in housework—a six-hour drop for women with pre-school children. By contrast, men’s paid and unpaid work hours changed very little over this period. Of course, households with different employment patterns do differ in the amount of time men and women spend on both paid and unpaid work; while full-time working mothers are employed for shorter hours than their partners (39 hours compared with 46 hours in 2003/04) they continue to do the bulk of the housework (14 hours compared with 7 hours for their male partner).

Paid and Unpaid Work 49 1991/92

30

Unpaid Work Hours

25

Wife: Housework hours Husband: Housework hours

20 15 10 5 0 0

10

20

25 30 35 Wives' Paid Work Hours

40

45

50

2003/04

30 25 Unpaid Work Hours

Wife: Housework hours Husband: Housework hours

20 15 10 5 0

0

10

20

25 30 35 Wives' Paid Work Hours

40

45

50

Source: Figures from Harkness (2008). Figure 3.6: Husbands and Wives: Unpaid Work Hours by Wives’ Paid Work Hours (British Household Panel Study)

Figure 3.6 shows the way the women’s and men’s unpaid housework hours are associated with the woman’s employment hours. The figures for 1991/92 and 2003/04 show that the pattern has not changed over time: women’s unpaid work diminishes rapidly, while men’s participation in unpaid work increases only very slightly.

50

Jacqueline Scott and Shirley Dex

Other analysis, using the longitudinal panel data that follow changes in individual patterns of paid and unpaid work over time (Gershuny, Bittman and Brice, 2005), points to a pattern of ‘lagged adaptation’. Thus although women decrease their unpaid work immediately and dramatically on returning to full-time work, their male partner’s corresponding increase in unpaid work is both delayed and somewhat unreliable. Gershuny et al find a similar pattern for men in the UK, the US and West Germany. Their findings raise the interesting question as to who is doing the work—whether unpaid chores are being left undone, or whether there has been an increase in the use of paid domestic services. There has been a clear growth in the demand for and supply of domestic services in recent years, but such services are often arranged and paid for by women and do little to ameliorate the gender imbalance in unpaid work. These findings also suggest that different national policies towards work–family balance have little effect on the gender division of labour in paid and unpaid work.

POLICY INITIATIVES CONCERNING WORK–FAMILY BALANCE

It is usually considered desirable to have greater equality between men and women in the labour market—and a lot of policy initiatives and anti-discrimination legislation have been orientated towards that goal. A related issue that is less visible in policy debate is whether steps should be taken to try to equalise women’s and men’s unpaid work loads in the home. In this section we consider the types of policy interventions used by governments which might, in principle, change unpaid work behaviour within households. As Lewis (2008) and others have pointed out, the policy regimes of many industrialised countries were designed and devised around the model of a male breadwinner family where the man worked full time and the women cared for the family and was not expected to be employed. This male breadwinner behaviour, in its pure sense, is hardly visible in industrialised countries of the twenty-first century because of the huge increases in women’s employment that have taken place. Although, as we showed earlier, many women do take time out of the labour force to have and to care for children, these periods have been getting successively shorter over recent generations (Macran, Joshi and Dex, 1996). For policy purposes the male breadwinner model still exists, albeit in a modified form. A common modification is for the male partner to be in paid work and working full-time hours, and the female partner to be in paid work but working part-time hours. A range of models that address work–family balance, together with the associated policies and example countries, is set out in Table 3.3. Policies have grown up in very different ways in different countries, and the logic underlying the policies can vary considerably from country to country. Moreover, some countries have adapted more quickly to the new models of family behaviour that have emerged, whereas other countries are slow to change (Lewis, 2006).

Paid and Unpaid Work 51 Table 3.3: Range of Models of Work–Family Balance Model/author

Description

Associated policies

Adult worker model family (Lewis, 2001). Comes in two forms:

Men and women are responsible for participating in the labour market.

a) supported

In work-benefits, tax Focus on getting lone parents and low credits acting as subsidy earners into work. to low paying employers.

Example countries

Stimulate provision of Model encouraged formal child-care services, in EU. possibly subsidised. UK since 1999, more so since 2003.

Tax relief or subsidy for child care if women in paid work. b) unsupported

Gender neutral, equality defined as sameness.

Earned income tax credits to make sure it is economic to work.

USA

No support for workers, except what is provided in the market. Market may provide cheap affordable domestic services and child care (eg, via high levels of immigration to offer low-wage work if the state colludes). Little support in leave or pay for childbearing, or income replacement while childbearing and child-rearing. Gender participation model, sometimes called the Nordic model, or ‘genderdifferentiated supported adult worker model’ (Hobson, 2004; Lewis, 2008)

Gender equality promoted, but makes allowances for difference.

Gender equality based on a women’s model of equality (Knijn, 2004)

All workers encouraged to reduce their weekly paid working hours to be part time.

Generous cash support for parental leave, services for child care and elderly dependents, but also for women to have extensive periods of leave (three years if two children born in quick succession) and rights to work part time until child is eight.

Sweden To a lesser extent in other Scandinavian countries. To a lesser extent in Germany.

The Netherlands

52

Jacqueline Scott and Shirley Dex

In principle there are two extremes that policy regimes can adopt: they can either support adults as paid workers, undifferentiated by gender, or they can acknowledge that men and women are likely to offer different levels of contributions to the labour market. No policy regime takes the extreme adult worker position, but the US comes pretty close to this, having only offered women rights to unpaid maternity leave since 1996. Scandinavian countries are often heralded as being more focused on providing equal opportunities to women and men, but their policies also allow women’s employment contribution to be different from men’s in having longer parental leave, and long periods of part-time work following childbirth. When policies allow or encourage women to behave differently in terms of their employment participation or their hours of work, gender differences in the home and in domestic contributions are tacitly endorsed. In order to link specific country policies with different time use patterns, Table 3.4 shows the mean time in minutes per day that men and women spend on different types of work and unpaid work, in the UK, the US, Sweden, the Netherlands and West Germany. These data are taken from time diaries of a longitudinal cross-national sample (Gershuny, 2000). ‘Paid work’ is contrasted with ‘core domestic work’ (referring to housework and cooking) and ‘other unpaid work’ (child care, shopping and odd jobs). Table 3.4 makes it clear that women in all these countries do a greater share of unpaid work than men. However, two other facts about the gender division of work are also worth noting. First, adding up women’s and men’s paid and unpaid work leads to near equality in the amounts of total work done by men and women (as in the Harkness figures cited earlier), or men doing slightly more total work

Table 3.4: Mean Time Spent per Day on Different Types of Work, in Minutes The Netherlands

UK

US

Sweden

West Germany

Core domestic work Men Women

29 188

28 177

33 182

56 143

11 238

Other unpaid work Men Women

84 124

83 111

97 142

117 146

84 132

Paid work Men Women

325 94

367 178

406 187

379 262

418 168

Total work Men Women

438 406

478 466

536 511

552 551

513 538

Source: Gershuny (2000: ch 7).

Paid and Unpaid Work 53 than women (the only exception being West Germany). Such figures suggest that claims of the ‘double burden’ (Hochschild, 1989) carried by women who are employed and still do the larger share of unpaid work may be exaggerated. Secondly, the average amounts of domestic work and paid work vary by country as by well as by gender, with relatively high total work hours in the US, Sweden and West Germany, and the lowest total work hours in the Netherlands. We suggested above that policies that make allowance for gender differences in employment practice are likely to reinforce gender differences in domestic work. In Sweden it is clear that women are spending more time on core domestic work than men, despite an explicit policy commitment to gender equality (Table 3.4). Nevertheless there is some evidence that policies supporting equality have some effect. The figures reported in Table 3.4 show Swedish men spending more time than their male counterparts elsewhere on core domestic work (56 minutes) and Swedish women spending the least time (143 minutes). However, even in Sweden, equality of unpaid domestic work seems an elusive goal. Policies restricting working time affect the potential time available to share in unpaid work. The EU has taken the initiative to direct members to limit weekly hours of paid work to a maximum of 48 hours per week. However, the UK allows opt-outs from this 1993 Directive1: it is not mandatory in the UK for all of its workers to comply with the 48-hour rule, and, not surprisingly, the UK now has the highest mean weekly paid working hours among men in Europe. Some countries have allowed parents the right to reduce their hours of work (for example Sweden and the Netherlands). But it is very largely women that use this ability to work part time. In 2003, the UK offered the right to request flexible working arrangements of their choice to parents of a child under the age of six. Employers were required to consider their request. This marked a new idea in UK industrial relations, moving away from voluntarism, but not as far as making this a statutory requirement. While such requests may be made by either parent, surveys show it is mainly women who make requests for and are offered flexible working arrangements (Palmer, 2004; Holt and Grainger, 2005). Is it possible for policies to influence gender shares of unpaid work? There is no compelling evidence that policies addressing work–family balance have had a significant or sizeable effect on the male and female share of unpaid work within households where they have been used. Policies however, can affect whether and

1 The European Union Working Time Directive was introduced in 1993 by the EU member states, with the aim of improving employment conditions. It was a legislative breakthrough, which changed employment law and set a maximum 48-hour working week. The 1993 Working Time Directive included Article 18, which allowed member states to opt out of the directive, and not apply the 48hour working week if a number of conditions were met. These included: workers must sign individual opt-out agreements, and must not suffer any penalty if they refuse to do so; employers must keep records of staff who work more than 48 hours a week, and make them available to the appropriate authorities. The opt-out from the Working Time Directive was not specific to the UK, but the UK was the only country within Europe to make widespread use of its provisions. In 2009 the EU Parliament voted to end the opt-out, but the UK government says it will appeal.

54

Jacqueline Scott and Shirley Dex

when women and men are in paid work during periods of childbirth and family formation. For example, ‘Daddy leave’ in Sweden (which is lost to the household if not taken by the male partner) has increased slowly from its initiation in 1995 rising to a 17 per cent take-up by 2003. Of course, even small increases in father involvement in child care might be highly beneficial in terms of child outcomes. Research in Britain (Dex and Ward, 2007) has suggested that fathers who took some parental leave around the birth of a child were more likely to read to a child aged three on a daily basis than fathers who did not take such leave (53 per cent as compared to 43 per cent). Also they found a correlation between fathers taking parental leave and three-year-old children having fewer behavioural and emotional problems. Studies have found women’s returns to work after childbirth have been sensitive to the conditions of their maternity or parental leave period (Ruhm, 1998; Brewer and Paull, 2006). In some cases, and with the most recent policies, it is perhaps too early to say that they have not achieved the goal of changing shares of unpaid work. But in the case of the policies based on gender difference, one could argue that they give to women with one hand, but take away with the other, as they reinforce the traditional gender role divide. This begs the question of whether the traditional gender role divide is what most women want. The case for preferences driving decisions about paid work has been argued by Hakim (2000), mainly in the context of the UK, although ‘preference theory’ is not without critics (Crompton, 2006). Hakim argues that it is possible to divide the female population into three groups according to their preferences: there are the career women who are focused on paid work; there are the homemakers who are focused on unpaid work and care; and, between these two, there is the adaptive group who will do paid work, but will give it up when it gets in the way of family commitments, since these have priority. The adaptive group is argued to be the majority of women among whom part-time work is very popular. It is certainly the case that part-time work is popular among some women, despite it being low paid and low skilled in some countries. There is also evidence in many countries that many women who are in full-time paid work would prefer to work fewer hours per week although this is also true of many men (OECD, 2001). The desire for flexibility in working hours and the extension of maternity leave rights has also been evident in Britain, especially among women (DTI, 2000). In expressing support for such options, women could be seen as embracing the difference approach to being paid workers, with lower hours, less attachment, less work experience, and, consequently, less pay and fewer career promotion prospects than men. Such policies facilitate an accommodation of gender inequality and a continuation of the unequal domestic division of unpaid work since they do not require the household boat to be rocked. Contemporary theorists emphasise that both men and women are ‘doing gender’, and both partners are living out traditional expectations of who does what. O’Brien’s review on ‘shared caring’ (2005) suggests that fathers do see the ‘good father’ role as including the role of primary breadwinner, but are happy for partners to contribute to the household income.

Paid and Unpaid Work 55 Counter to this claim that a redistribution of unpaid work is not what women want, others would point out that the so-called ‘choices’ that parents make are still being made on a playing field that is not level or equal between genders. There are a range of other policies that support the (higher-paid) male partner working longer paid hours than the female, and there is still the unequal wage rate issue. Nonetheless, we cannot discount people’s expressed preferences. Fathers are largely content with the hours they work, even when their work hours are as much as 60 hours per week (O’Brien, 2005). Mothers like part-time paid work; they like flexibility in their working hours; they are generally happy with the care policies that acknowledge that women are different and treat them differently. It seems unlikely that equality in either paid or unpaid work will come from such preferences.

DOES GENDER IMBALANCE IN PAID AND UNPAID WORK MATTER?

If women do less paid work outside the home than men, then it seems only equitable that they should do more unpaid work in the home than men. In principle, the female partner could do more paid work than the man, but the relative wage rates are against this choice and in favour of women’s specialisation in home work (Becker, 1991). Men will be able to earn more per hour, on average, than women. So it is more efficient for the man to work, and thus, he accumulates more human capital which will bring him higher wages in future. But this reinforces the unequal wage rates for men and women and locks women into unpaid home work. Does this matter? If the couple have committed themselves to living together as a unit, then they both stand to gain financially by this gender specialisation. In the past, couples were happy to do this, but times are changing. It is now seen as riskier for the woman to compromise her earning potential. She needs to maintain human capital (in terms of work experience and training) to cope with future uncertainties, such as unemployment or divorce (see Fisher and Low, chapter eleven, this volume). Moreover, the traditional female career pattern has exerted a heavily penalty on older women who are reliant on state pensions (see Price, chapter twelve, this volume). It is also a problem in countries like the US where health insurance is tied to employment. So, what can be done and what should be done to address the gender imbalance? Is it possible to change the wage ratio in order to make a more level playing field for men’s and women’s intra-household decision-making? Is it desirable to coerce men to do more of the unpaid work and family care, even if that were possible? It is sometimes suggested that compulsion to care goes against the inherent meaning of the caring activity. It is claimed that one cannot force someone to be responsible and attentive in a competent manner. Caring usually requires some feeling of empathy to shape appropriate actions. The most that is possible is to give people the choice to care and to provide favourable conditions in which they can

56

Jacqueline Scott and Shirley Dex

exercise such choice. One could argue, however, that many women are constrained to care because of lack of alternatives, whether nursery places or a partner who is willing and able to share in the caring. There are positive signs, however, that in practice men have increased their contribution to child care more than to other core domestic duties. Policies such as ‘daddy leave’ are surely to be welcomed, in so far as they help support men’s wish to take a more active fathering role. A more problematic and contentious issue is how to address the unequal wage rates of men and women. State policy attempts to change the female-to-male wage ratio in order to achieve changes in the shares of unpaid work are not in evidence. The idea of giving wages for housework has been suggested and discussed, but never implemented (for example Young and Halsey, 1995). In the US in the 1970s, the possibility of crediting homemakers with social security contributions was discussed but not adopted. Women’s behaviour in entering the labour market in large numbers has made redundant any policy interest in such ideas. Cash for parental care is a policy that has many examples, including parental leave. More recent examples used in Finland and Norway offer parents the choice between cash to care for their own children at home or a subsidised place in formal child care. Such policies have been popular among lower-paid women, who are the ones who have taken the cash and stayed at home rather than choosing the child-care places. These policies are criticised by feminists, who think that women will only be emancipated through employment. But none of these policies have tried to manipulate women’s wages in the labour market. It is not likely to be possible suddenly to change the amounts of human capital that are embedded in individuals’ wage rates, such as the differing amounts of paid work experience that men and women have. But legislating for equal pay for equivalent work is a policy that starts to tackle the issue, so long as it is actually implemented in workplaces. Similarly, gender pay audits and pay reviews, as well as enforced monitoring of pay and equal opportunities, can assist in making sure women do not fall behind when they are in paid jobs. Unfortunately, the current move in the UK to ‘reflexive regulation’ adopted by the Discrimination Law Review consultation document is unlikely to be effective in achieving greater pay equality, especially outside the public sector, at a time when collective bargaining has diminished and when there are no appropriate institutional mechanisms for carrying through equality change (Deakin and McLaughlin, 2008). Deakin and McLaughlin’s pessimistic appraisal of the likely success of the current ‘encouragement’ of redressing gender pay inequalities is born out by cross-national data. The result of several decades or more of trying to achieve gender wage equality is that both raw and adjusted-for-work-experience female-to-male wage ratios remain resistant to equality in nearly all countries. One approach to raising the wages of partnered women relative to men would be to tax partnered men’s wages sufficiently to give women in paid work a sizeable tax credit to boost their hourly wage rate to the same level as their partner’s aftertax hourly rate. This policy could, in principle, equalise wage rates and eliminate the incentive for the female partner to be the person who did more of the unpaid

Paid and Unpaid Work 57 work. Whether equality in wage rates would be sufficient to get the women to do more paid work is not clear, since there is still considerable evidence that women like caring (Houston and Marks, 2005) and some evidence that men cannot do ironing (for example, Moir and Moir, 2000). Such a policy initiative, however, is not going to happen. No government would see such an aggressive approach for reducing gender inequalities in division of labour within couples as either feasible or desirable. So we are back to token or symbolic steps to try to encourage men to do a more equitable share of domestic and care work. Tokens and symbolic gestures, however, do matter, and signals from the state encouraging greater male participation in unpaid work could help advance the slow pace of gender convergence. There might also be useful steps that parents and schools could take to help the next generation reduce gender inequalities. According to data from the youth survey of the British Household Panel, the contribution that boys and girls aged 11–15 are making to housework show clear gender differences—with more than a third of boys doing little or no housework compared with one fifth of girls (Harkness, 2008). But whether such a difference is increasing or diminishing only time will tell. It may, however, be more feasible to tackle gender inequalities in youth through educational and parental encouragement than to change ingrained gender inequalities in paid and unpaid work among the adult population.

BIBLIOGRAPHY Becker, G (1991) A Treatise on the Family (Cambridge, MA, Harvard University Press). Brewer, M and Paull, G (2006) Newborns and New Schools: Critical Times in Women’s Employment (London, Department for Work and Pensions). Crompton, R (2006) Employment and the Family (Cambridge, Cambridge University Press). Deakin, S and McLaughlin, C (2008) ‘The Regulation of Women’s Pay: From Individual Rights to Reflexive Law?’ in J Scott, S Dex, and H Joshi (eds), Changing Patterns of Women’s Employment over 25 Years (Cheltenham, Edward Elgar). Del Boca, D and Wetzels, C (2007) Social Policies, Labour Markets and Motherhood: A Comparative Analysis of European Countries (Cambridge, Cambridge University Press). Dex, S (forthcoming) ‘Policy Interventions to Equalize Men’s and Women’s Time Spent in Unpaid Work: Are they Possible and Realistic?’ in J Treas and S Drobnic (eds), Dividing the Domestic: Men, Women and Household Work in Cross-National Perspective (Stanford, CA, Stanford University Press). Dex, S and Ward, K (2007) Parental Care and Employment in Early Childhood: Analysis of the Millennium Cohort Study (MCS) Sweeps 1 and 2 (Manchester, Equal Opportunities Commission Report). Dex, S, Ward, K and Joshi, H (2008) ‘Changes in Women’s Occupations and Occupational Mobility over 25 years’ in J Scott, S Dex and H Joshi (eds), Changing Patterns of Women’s Employment Over 25 Years (Cheltenham, Edward Elgar). DTI (Department of Trade and Industry) (2000) Work and Parents: Competitiveness and Choice, a Research Review (London, DTI).

58

Jacqueline Scott and Shirley Dex

Gallie, D (2000) ‘Labour Force Change’, in A Halsey with J Webb (eds), British Social Trends, 3rd edn (Houndsmill, Macmillan Press Ltd). Gershuny, J (2000) Changing Times: Work and Leisure in Post Industrial Society (Oxford, Oxford University Press). —— (2004) ‘Time through the Life-course in the Family’ in J Scott, J Treas and M Richards (eds), The Blackwell Companion to Sociology of Families (Oxford, Blackwell). Gershuny, J, Bittman, M and Brice, J (2005) ‘Exit, Voice and Suffering: Do Couples Adapt to Changing Employment Patterns?’ 67 Journal of Marriage and Family 656. Hakim, C (2000) Work–Lifestyle Choices in the 21st Century: Preference Theory (Oxford, Oxford University Press). Harkness, S (2008) ‘The Household Division of Labour: Changes in Families Allocation of Paid and Unpaid Work’ in J Scott, S Dex and H Joshi (eds), Changing Patterns of Women’s Employment Over 25 Years (Cheltenham, Edward Elgar). Hobson, B (2004) ‘The Individualised Worker, the Gender Participatory and the Gender Equity Models in Sweden’ 3 Social Policy and Society 75. Hochschild, A (1989) The Second Shift, Working Parents and the Revolution at Home (London, Piatkus). Holt, H and Grainger, H (2005) Results of the Second Flexible Working Employee Survey, DTI Employment Relations Research Series No 39 (London, Department of Trade and Industry). Houston, DM and Marks, G (2005) ‘Working, Caring and Sharing: Work–Life Dilemmas in Early Motherhood’ in D Houston (ed), Work–Life Balance in the Twenty-first Century (London, Palgrave Macmillan). Joshi, H and Paci, P (1998) Unequal Pay for Women and Men: Evidence from the British Birth Cohort Studies (Cambridge, MA: MIT Press). Joshi, H (1989) ‘The Changing Form of Women’s Economic Dependency’ in H Joshi (ed), The Changing Population of Britain (Oxford, Basil Blackwell). Knijn, T (2004) ‘Challenges and Risks of Individualisation in the Netherlands’ 2 Social Policy and Society 57. Kohler, H-P, Billari, F and Ortega, JA (2006) ‘Low Fertility in Europe: Causes, Implications and Policy Options’ in F Harris (ed), The Baby Bust: Who will do the Work? Who will Pay the Taxes? (Lanham, MD, Rowman & Littlefield Publishers). Lewis, J (2001) ‘The Decline of the Male Breadwinner Model: The Implications of Work and Care’ 8 Social Politics 152. —— (2006) ‘Men, Women, Work, Care and Policies’ 16 Journal of European Social Policy 387. —— (2008) ‘Work–Family Balance Policies: Issues and Development in the UK 1997–2005 in Comparative Perspective’ in J Scott, S Dex and H Joshi (eds), Changing Patterns of Women’s Employment Over 25 Years (Cheltenham, Edward Elgar). Macran, S, Joshi, H and Dex, S (1996) ‘Employment after childbearing: A survival analysis’ 10 Work Employment and Society 273. Moir, A and Moir, B (2002) Why Men Don’t Iron: The Science of Gender Studies (London, Citadel, Kensington Publishing Corps). O’Brien, M (2005) Shared Caring: Bringing Fathers into the Frame (London, Equal Opportunities Commission). OECD (Organization for Economic Cooperation and Development) (2001) Employment Outlook 2001 (Paris, OECD).

Paid and Unpaid Work 59 Palmer, T (2004) Results of the First Flexible Working Employee Survey, DTI Employment Relations Occasional Papers URN 04/703, www.dti.gov.uk/er/emar. Purcell, K and Elias, P (2008) ‘Achieving Equality in the Knowledge Economy’ in J Scott, S Dex and H Joshi (eds), Changing Patterns of Women’s Employment Over 25 Years (Cheltenham, Edward Elgar). Royal Commission on Population (1949) Report (London, HMSO). Ruhm, C (1998) ‘The Economic Consequences of Parental Leave Mandates’ 113 Quarterly Journal of Economics 285. Scott, J (2008) ‘Changing Gender Role Attitudes’ in J Scott, S Dex and H Joshi (eds), Changing Patterns of Women’s Employment Over 25 Years (Cheltenham, Edward Elgar). WEU (Women and Equality Unit) (2004) Women and Men in the Workplace, www. womenandequalityunit.gov.uk/research/gender_briefing_nov04.doc. Young, M and Halsey, AH (1995) Family and Community Socialism (London, Institute for Public Policy Research).

4 Managing Money in Intimate Relationships Similarities and Differences between Cohabiting and Married Couples CAROLYN VOGLER

INTRODUCTION

T

his chapter aims to explore what we can learn from two recent, largescale, nationally representative surveys, about the similarities and differences between different subcategories of cohabiting and married couples, in terms of the financial context of their relationships and how they handle money within them.1 The latter is important, not only because it has serious implications for the distribution of property and possessions when relationships end (Tennant, Taylor and Lewis, 2006), but also because it has important implications for the well-being and satisfaction of each individual during the life of a relationship (Vogler, Lyonette and Wiggins, 2008). There is now a rich sociological literature showing how the different ways in which couples manage money can be seen as a tangible expression of how they resolve the tensions at the heart of all intimate relationships between, on the one hand, individual autonomy versus commitment to the welfare of the couple as a collective unit, and, on the other, equality versus inequalities in power and living standards between individuals within the same relationship (Blumstein and Schwartz, 1983; Pahl, 1989; Burgoyne, 1990; Treas, 1993; Vogler and Pahl, 1993; Fleming, 1997; Singh, 1997; Nyman, 1999; Treas and Widmer, 2000; Elizabeth, 2001; Rake and Jayatilaka, 2002; Heimdal and Houseknecht, 2003). Traditionally, of course, these polarities have been institutionally rooted in the marriage contract and the labour market. While

1 I am grateful to the Economic and Social Research Council (ESRC) for providing funding to undertake this research (ESRC grant no R000239727). I am also grateful to Rosemary Crompton for acting as the principal applicant, and to the ESRC’s anonymous referees for insightful comments and very helpful suggestions.

62 Carolyn Vogler the marriage contract requires spouses to support each other financially, both during marriage and after it ends (Barlow et al, 2005), we enter the labour market as individuals who are in some sense deemed to ‘own’ the money we have earned: it is ‘ours’ and we are seen as having a legitimate right to do what we like with it (Burgoyne 1990; Burgoyne and Lewis 1994; Singh, 1997; Nyman and Reinikainen 2002; Rake and Jayatilaka, 2002). In the past, these tensions were resolved in a highly gendered way. Men were expected to act as breadwinners for the whole family and were therefore seen as having a legitimate right to more control over, and access to, money for their own personal use. Wives, on the other hand, were constructed as economically dependent on husbands and primarily responsible for unpaid caring and household tasks, regardless of how much they actually earned (Zelizer, 1989, 1994). Today, of course, this has changed dramatically. Both men and women now participate in the labour market as ‘individuals’ with their ‘own’ incomes, and intimate relationships are almost invariably regarded as partnerships between equals, in which all resources are ideally shared equally, regardless of who contributes what to the household. At the same time, however, men still earn more than women, while women are still seen as responsible for unpaid caring and household tasks (see Scott and Dex, chapter three, this volume). In these circumstances, the problem couples face is that individual autonomy and equality are inevitably in tension with each other, because equality requires higher earners to give up a degree of individual autonomy and control over their ‘own’ income in order to redistribute equally within the couple. The intra-household economy can therefore be seen as a crucially important dimension of intimate relationships, sitting at the interface between the couple and the wider society, mediating the extent to which gender inequalities in the labour market are transposed into inequalities in access to and control over money within the relationship (Vogler and Pahl, 1994). One fundamental question that arises in this context is whether different forms of money management are adopted in different types of relationships. In recent years, the dramatic increase in cohabitation outside marriage has led to an important debate about the similarities and differences between cohabiting and married couples. But, despite its obvious relevance, given the different legal and financial ties between married and cohabiting couples, one issue that has seldom been addressed in the context of this debate is the financial dimension of intimate relationships. While qualitative studies indicate that in many ways cohabiting couples tend to be very similar to their married counterparts, especially in relation to commitment, attitudes, values and the rhythm of their everyday lives together (McRae, 1993; Lewis, 2001; Jamieson et al, 2002; Eekelaar and Maclean, 2003), two large-scale international studies indicate that there may nevertheless be some very big differences in both the financial context of their relationships and how they handle money within them (Treas and Widmer, 2000; Heimdal and Houseknecht, 2003). In the UK, these have important implications for policy when couples split up.

Managing Money in Intimate Relationships 63 Drawing on quantitative, nationally representative British data from two annual sweeps of the International Social Survey Programme (ISSP),2 in 1994 and 2002, for modules focusing on ‘Family and Changing Gender Roles’, this chapter seeks to explore the similarities and differences between cohabiting couples and their married counterparts in the same age range and family situations, by addressing four key questions. First, are there any differences between different subcategories of married and cohabiting couples in terms of their background characteristics and socio-economic circumstances? Secondly, how far do different subcategories of cohabiting and married couples manage money differently from each other, after controlling for a range of demographic, socio-economic, and ideological/ discursive factors which have been found to be important predictors of how couples manage money in previous studies, such as age, the presence of children, both partners’ relative economic contributions to the household, ideologies/discourses of breadwinning, and social class? Or how far, on the other hand, are cohabiting couples just like married couples without an official licence? Thirdly, how far are cohabiting couples now coming to be more egalitarian than their married counterparts in terms of (1) access to money for personal spending and saving and (2) who has the final say over important expenditure decisions? And, finally, although there is little previous research in this area, how far is power over expenditure decisions related to satisfaction with the relationship and happiness with life in general? In order to put the analysis in context, I begin with a brief discussion of existing research on patterns of money management and how they have changed over time, before focusing on this specific data and the four questions raised above. Finally, I conclude by highlighting the main findings and drawing out their implications for policy. SYSTEMS OF MONEY MANAGEMENT

When couples first set up home together they inevitably have to establish some way of managing the money coming into the household from earnings, state benefits and so on, in order to meet bills for basic living expenses such as mortgage/ rent, council tax, fuel, food and so on, as well as personal spending money (Pahl, 1989). However, given that couples usually earn different amounts and that one or both partners may have no income, except possibly state benefits, this is not necessarily straightforward. If couples have different ideas about money or different assumptions about who should pay for what and how much it is legitimate to spend on oneself, without consulting the other partner first, it can all too easily lead to disagreement, underlying resentment and sometimes open conflict.

2 The ISSP is a voluntary grouping of study teams in 38 nations each of which agrees to run a short annual self-completion survey containing an agreed set of questions asked of a probabilitybased nationwide sample of adults aged 18 and over (see Davis and Jowell, 1989; also www.issp.org. uk/info.htm).

64 Carolyn Vogler Organising money can also be difficult because in Western societies the financial aspects of intimate relationships tend to be a ‘taboo subject’, which couples dislike talking about, even in private, particularly during the course of an ongoing relationship (Blumstein and Schwartz, 1983; Hertz, 1986, 1992; Rake and Jayatilaka, 2002). Qualitative studies indicate that there is very little explicit discussion or negotiation about how to manage money: couples frequently say they never sat down and tried to work out rationally the best system for them, instead they often claim they just did ‘what came naturally’ (Pahl, 1989; Tennant, Taylor and Lewis, 2006). This supports the finding by Douglas, Pearce and Woodward (chapter seven, this volume) that patterns of money management tend to emerge more by default or force of personality, than, as economists have traditionally assumed (Becker, 1993), by rational consideration and open negotiation. Not surprisingly, this tends to result in partners’ having different assumptions and expectations from each other when relationships end (Tennant, Taylor and Lewis, 2006). In the early 1980s, Pahl (1989) devised a typology of the main ways in which couples organise money, based on how couples manage money on a day-to-day basis, rather than how they exercise strategic control (or power) over it. This is because, as in a company or a work environment, the person who manages money on a daily basis may be different from the person with ultimate control over strategic financial decision-making. Traditionally, wives have been more likely to manage money as part of their domestic responsibilities for shopping, while control has more often been a male prerogative, associated with the breadwinner or primary earner status (Edwards, 1981; Blumstein and Schwartz, 1983; Pahl, 1989; Vogler and Pahl, 1993). The typology is divided into two subsections depending on the extent to which money is either constructed as jointly owned by both partners, so that individual autonomy is to some degree subordinated to the needs of the couple as a collective unit; or whether couples prioritise individual freedom and operate largely as two separate, autonomous individuals, each with their own individually owned money, so that neither is completely financially dependent on the other (Treas, 1993; Treas and Widmer, 2000; Elizabeth, 2001; Nyman and Reinikainen, 2002). Systems in which couples operate more or less as single economic units include: — the female whole wage system in which women manage all the money except the man’s personal spending money; — the male whole wage/housekeeping allowance system in which men either manage all the money (which may leave non-earning wives with no personal spending money) or men manage most of the money, except for the woman’s housekeeping allowance; — the joint pooling system in which couples pool all the money, usually in a joint bank account and in theory manage it jointly, each taking money out as needed.

Managing Money in Intimate Relationships 65 Individualised systems in which couples operate largely as two separate autonomous economic units include: — the partial pool in which couples pool some of their income to pay for collective expenditure and keep the rest separate to spend as they choose, without having to discuss it with the other partner; — finally, the independent management system in which both partners have their own independent incomes from earnings or state benefits which they keep completely separate, and each partner has responsibility for different items of household expenditure. While the female whole wage and housekeeping allowance systems involve separate spheres of expenditure for men and women based on a discourse of breadwinning in which men are constructed as the main breadwinners, with a legitimate right to more control over money and more money for their own use, the joint pool is less internally structured and in theory offers greater scope for discussion and negotiation over the use of pooled money (Shove, 1993). While pooling is in theory based on a discourse of equality in which all money is defined as jointly owned and both partners have equal access to and control over money, this tends to be achieved in practice only when both partners are in full-time employment and also see themselves as co-providers, equally responsible for providing the income. When the male partner is constructed as the main breadwinner, equal outcomes tend to be undermined by a more implicit and pervasive discourse of earner control, legitimating inequalities in personal spending money, favouring men (Pahl, 1989; Burgoyne, 1990; Shove, 1993; Vogler and Pahl, 1993, 1994; Fleming, 1997; Singh, 1997; Elizabeth, 2001). So far there has been little research on the two privatised systems, although the little that does exist indicates that couples using the independent management system attempt to establish both equality and individual autonomy by constructing all money as individually owned and by defining equality not as equal access to and control over money, as in the joint pool, but as making equal contributions towards collective expenditure (going 50:50) regardless of income (Brines and Joyner, 1995; Singh and Lindsay, 1996; Elizabeth, 2001; Nyman, 2003). While keeping money separate thus avoids problems of financial dependency, it does not necessarily produce equal outcomes, because, if one partner earns more than the other, the higher earner ends up with more financial power in the relationship and more money for his or her own use (Pahl, 2005; Burgoyne et al, 2006). In a series of qualitative interviews with couples in New Zealand, Fleming (1997) found that partial pooling couples often seemed to have a foot in each camp. While they saw the household unit as important and liked the idea of pooling, they were often not fully committed to it in practice and sometimes disagreed over the extent to which money was jointly or separately owned. He concluded that the partial pool could therefore be thought of as a hybrid between systems in which households operate as single economic units and those in which money is kept completely separate,

66 Carolyn Vogler acting as a way of accommodating the two conflicting principles of jointness and autonomy.

Changing Patterns of Money Management among all British Heterosexual Couples In order to provide a rough indication of the use of Pahl’s (1989) systems among British (heterosexual) couples as a whole (married and cohabiting), as well as how this has changed over time, Table 4.1 compares the use of different allocative systems at three different points in time: 1987 (Vogler and Pahl, 1993, 1994), 1994 and 2002 (Vogler, Brockmann and Wiggins, 2006).3 As can be seen in the table, the most commonly used system at all three points in time was clearly the joint pool, chosen by approximately half of respondents. The table also shows a slight decline in the use of the two traditional systems based on the male breadwinner model of gender (the female whole wage and housekeeping allowance systems) and a slight increase in the use of the partial pool, so that by 2002 it had overtaken the female whole wage system as the second most frequently used system. In the samples as a whole, there were no significant differences between male and female respondents’ answers. Table 4.1 also shows that the female whole wage system was associated with the lowest average standardised household income, while the partial pool and the independent management systems were associated with the highest incomes, with the housekeeping allowance and joint pooling systems falling between the two extremes. This supports previous studies showing that female partners are most likely to manage money in low-income households where there is insufficient money to meet the bills and the task is likely to be a chore or a burden rather than a source of power, while male partners, on the other hand, are more likely to manage money when income is high enough to allow for discretionary spending and it is possible for real decisions to be made about how to spend money (Pahl, 1989; Vogler and Pahl, 1993, 1994; Singh, 1997; Goode, Callender and Lister, 1998; Rake and Jayatilaka, 2002). Given that they presuppose two separate sources of income, the partial pool and the independent management system were, not surprisingly, most likely to be used by the highest-income couples who, as we shall see later, were more likely than other couples to both be in full-time employment. Previous research indicates that allocative systems are largely the result of two inter-related factors: first, the relative economic resources each partner

3 The 1987 data consist of a sub-sample of 1,211 individuals living in couple relationships, who were initially included as part of the Social Change and Economic Life Initiative. This was based on interviews with a random sample of 6,000 adults aged 20–60, living in six towns in Britain in 1986 (Gallie, 1988). In terms of their economic and demographic characteristics the 1,211 individuals broadly corresponded to an national sample (Vogler and Pahl, 1994).

Managing Money in Intimate Relationships 67 Table 4.1: Proportion of Opposite-sex Couples in Britain using Different Household Allocative Systems in 1987, 1994 and 2002, together with Mean Standardised Household Income in 2002 (male and female respondents in couple relationships) 1987

1994

2002

2002

Allocative Allocative Allocative Mean Systems (All) Systems (All) Systems (All) Standardised Household Income % Female whole wage 26 Housekeeping allowance 22 Joint pool 50 Partial pool (not included) Independent management 2 Total (%) Respondents (N) Significance level

100 (1,211)

% 14 10 54 13 9 100 (686)

% 10 8 55 17 10

£ per year 16,861 19,879 24,398 30,799 29,807

100 (1,308) p < 0.001

Note: Income was standardised using the McClelland Equivalence Scale, according to the number of adults and children in the household, obtained from the British Household Panel Survey User Manual Volume A (Taylor, 2005: table 29). No significant differences between men’s and women’s answers in 1987, 1994, or 2002. Source: 1987 figures from the Social Change and Economic Life Initiative (Vogler and Pahl, 1994), and 1994 and 2002 figures from the ISSP (Vogler, Brockmann and Wiggins, 2006).

contributes to the household (as measured by household employment status rather than income, because not all couples are economically active) and, secondly, cultural ideologies/discourses of gender, particularly those of male breadwinning versus newly emerging discourses of co-provisioning (Vogler and Pahl, 1993, 1994; Ellingsaeter, 1998). One factor which has seldom been included in British research on this topic is the type of relationship couples establish. To see how far different subcategories of cohabiting and married couples organise money differently after controlling for selection effects, we therefore need to turn to our own data.

DATA USED IN THE ANALYSIS

Our own comparison of different subcategories of cohabiting and married couples is based on British data from two cross-sectional, national surveys of ‘Family and Changing Gender Roles,’ undertaken by the UK’s National Centre For Social Research, as part of the 1994 and 2002 ISSP. Both surveys were based on nationally representative samples of the adult population (aged 18 and over) living in Great

68 Carolyn Vogler Britain (that is, England, Wales and Scotland, south of the Caledonian Canal). Each respondent participated in a face-to-face interview and was left with a selfcompletion questionnaire to fill in after the interview, which was returned by post. The 1994 survey yielded 688 male and female respondents aged 18 and over, living in married and opposite-sex cohabiting relationships, of whom 9 per cent were cohabiting and 91 per cent were married. The 2002 survey yielded 1,308 similar respondents, of whom 13 per cent were cohabiting and 87 per cent were married. Comparison with the 2002 General Household Survey (GHS) indicated that the level of cohabitation in the ISSP was broadly representative of the population as a whole (for more details, see General Household Survey, 2004; Vogler, Brockmann and Wiggins, 2006, 2008; Vogler, Lyonette and Wiggins, 2008).

Identifying Subcategories of Cohabiting and Married Respondents Our first objective was to identify the three main subcategories of cohabiting respondents identified by Kiernan and Estaugh (1993) in their analysis of the 1989 GHS, namely, ‘never-married childless cohabitants’, ‘never-married cohabiting parents’, and ‘post-marital cohabitants’, together with their married counterparts in the same age range and family situations. Unfortunately, we were unable to use the 2002 ISSP for the entire analysis because, unlike the earlier 1994 survey, it did not include questions on marital or partnership histories and thus did not enable us to identify post-marital cohabiting and remarried respondents separately. For this, we had to return to the earlier 1994 data. Our sub-samples of ‘never-married childless cohabitants’ and ‘never-married cohabitants with children’ were drawn from the 2002 data. Since national statistics (Summerfield and Babb, 2003; Walker, 2003), indicate that never-married cohabitants are overwhelmingly young (under 35 years old), both samples consisted of all respondents under 35 years old, who were cohabiting with a partner at the time of the 2002 survey. They were then matched with a sample of married couples in the same age range. This yielded 306 male and female respondents under 35 years old, over a third (36 per cent) of whom were cohabiting while just under two thirds (64 per cent) were married. Both groups were then divided into those with and without children aged 16 or under living with them. This generated four subcategories which we refer to as ‘childless cohabiting respondents’, ‘cohabiting parents’, ‘childless married respondents’ and ‘married parents,’ respectively. It is important to note that the category of cohabiting parents needs to be treated with extreme caution, partly because of the small number of respondents in this category (38), but also because the 1994 data indicate that in 1994, approximately one quarter of cohabiting parents under 35 were post-marital cohabitants, who as we see later, tend to be very different from never-married cohabiting parents. Our findings about cohabiting parents therefore need to be seen as indicative rather than conclusive, although it was reassuring to find that they were broadly consistent with the findings of previous research (Kiernan and Estaugh, 1993;

Managing Money in Intimate Relationships 69 McRae, 1993, 1999; Pickford, 1999; Smart and Stevens, 2000; Lewis, 2001; Lewis, Papacosta and Warin, 2002; Ermisch, 2005). Our sub-sample of ‘post-marital cohabitants’ consisted of 27 post-marital cohabitants aged between 26 and 56 in the 1994 survey, who were matched with 92 remarried respondents in the same age range. We refer to them as ‘post-marital cohabitants’ and ‘remarried respondents’, respectively. Both groups include those with and without children under 16 living with them. Given the small number of post-marital cohabitants, it was not possible to subdivide them further, but age and the presence of children are controlled for in all our later multivariate analysis. It needs to be stressed that since the three subcategories of cohabiting and married respondents were drawn from different data sets and constructed on the basis of different criteria, they are not directly comparable with each other. However, since the aim is to compare cohabiting and married respondents within sub-samples rather than between them, we can still make meaningful inferences, although the results clearly need to be treated as tentative. Tables 4.2 and 4.3 show the proportion of respondents living in each type of relationship at the time of the respective surveys, together with their average age in years. As can be seen in the tables, the data are consistent with previous studies showing that cohabiting respondents tend to be younger than their Table 4.2: Relationship Status and Mean Age of Partnered Respondents (aged under 35) in Britain, in 2002 Relationship Status

Childless married respondents Childless cohabiting respondents Married parents Cohabiting parents Total (%) Respondents (N)

%

20 24 44 12

Mean Age in Years Mena

Womenb

29.2 27.6* 30.3 29.3

29.6 26.1*** 30.6 27.9***

100 (306)

Notes: *p < 0.1; **p < 0.05; ***p < 0.01. a. Overall differences between men ANOVA F = 3.5; d.f. = 3; p < 0.05. T tests showed childless cohabiting men were marginally younger than childless married men (t = −1.7; d.f = 126; p < 0.1) but the difference between married and cohabiting fathers was not significant. b. Overall differences between women ANOVA F = 14.2; d.f. = 3; p < 0.001. T tests showed childless cohabiting women were significantly younger than childless married women (t = −4.073; d.f = 171; p < 0.001), and cohabiting mothers were significantly younger than married mothers (t = −3.087; d.f. = 171; p < 0.01). No significant differences between men and women within relationship categories. Source: ISSP 2002, Vogler, Brockmann and Wiggins (2008).

70 Carolyn Vogler Table 4.3: Relationship Status and Mean Age of Repartnered Post-marital Respondents (aged between 26 and 56), in Britain in 1994 Relationship status

Post-marital married respondents Post-marital cohabiting respondents Total (%) Respondents (N)

%

77 23

Mean age in years Mena

Womenb

39.1 39.2

41.8 37.2

100 (119)

Notes: *p < 0.1; **p < 0.05; ***p < 0.01. a Age

differences between men not significant. Age differences between women not significant. No significant differences between men and women within relationship categories. b

Source: ISSP 1994, Vogler, Brockmann and Wiggins (2008).

married counterparts, although in the case of cohabiting fathers and post-marital cohabitants, the differences did not reach statistical significance (Kiernan and Estaugh, 1993; Ermisch, 2005). The main findings will be presented in four sections corresponding to the research questions specified above. The first two sections draw on both the 1994 and 2002 data to compare the three different subcategories of cohabiting couples with their married counterparts in the same age range and family situations. In contrast, the third and fourth sections attempt to boost the number of cohabiting couples available for analysis in the same data set by using all cohabiting and married couples in the 2002 data, without separating out couples over 35 years old who may have been married before. What sort of people, then, cohabit rather than marry, and how do they compare with married people in terms of their background characteristics and socio-economic circumstances?

Background Characteristics and Socio-economic Circumstances Tables 4.4 and 4.5 show the main differences between different subcategories of cohabiting and married couples in terms of education, employment status, mean standardised household income, housing tenure and social class. Our findings broadly support Kiernan and Estaugh’s (1993) analysis of the 1989 GHS, showing that, in terms of education, income and social class, childless and post-marital cohabitants were quite similar to their married counterparts, although they were less likely to be living in owner-occupied accommodation and more likely to be living in households in which both partners were in full-time

Managing Money in Intimate Relationships 71 Table 4.4: Socio-economic Characteristics of Cohabiting and Married Respondents (aged under 35) in Britain, in 2002 Childless Childless Married Married Cohabiting Parents Respondents Respondents A. Educational qualifications Men % with A-levels and above % with O-levels % with no formal qualifications Women % with A-levels and above % with O-levels % with no formal qualifications B. Household Employment Status % both employed full-time % man FT woman PT % man FT woman not employed % no earners in the household C. Mean Standardised Household Income (£/year) D. Housing tenure % owner occupiers % renting form local authority % renting privately E. Social Class Women % managerial/professional % intermediate class % working class

Cohabiting Parents

93 3 4 n.s.

81 16 3

53 36 43 43 4 21 χ2 = 4.794; d.f. = 2; p < 0.1

81 16 3 n.s.

71 29 –

52 32 44 54 4 14 χ2 = 4.769; d.f. = 2; p < 0.1

74 12 2 2 n.s.

85 4 4 1

18 18 34 26 34 1 6 18 χ2 = 19.331; d.f. = 4; p < 0.05

39,249 n.s.

36,999

22,626 n.s.

19,522

87 65 77 46 3 4 15 43 10 31 7 11 χ2 = 8.84; d.f. = 2; p < 0.05 χ2 = 15.1; d.f.= 2; p < 0.001

50 29 21 n.s.

56 21 23

40 24 27 9 33 67 χ2 = 7.771; d.f. = 2; p < 0.05

Notes: Total household income was standardised by the number of adults and children in the household using the McClement’s Equivalence Scale (Taylor, 2005). Class was measured on the Goldthorpe three-class model (Goldthorpe, 1987). Source: ISSP 2002, Vogler, Brockmann and Wiggins (2008).

72 Carolyn Vogler Table 4.5: Differences in Socio-economic Characteristics between Post-marital Cohabiting and Remarried Respondents (aged between 26 and 56) in Britain, in 1994 Remarried Respondents A. Household Employment Status % both employed full-time % man FT; woman PT % man FT; woman not employed % other B. Housing Tenure % owner occupiers % renting from local authority % renting privately

Post-marital Cohabiting Respondents

29 61 24 11 18 11 30 18 χ2 = 9.73; d.f. = 3; p < 0.05 75 54 17 39 9 7 χ2 = 6.538; d.f. = 2; p < 0.05

Source: ISSP 1994, Vogler, Brockmann and Wiggins (2008).

employment. As Kiernan and Estaugh (1993) point out, in the case of young childless cohabitants, these differences may be at least partly because they are younger than their married counterparts and thus likely to be at an earlier stage in their partnership and housing histories. Turning to those with children, however, the picture was very different: cohabiting parents stood out in being socio-economically disadvantaged compared with their married counterparts. They were more likely than married parents to have no formal qualifications, to be living in social housing (43 per cent and 15 per cent, respectively) and to have no earners in the household (18 per cent and 6 per cent, respectively). This was mainly because cohabiting fathers were only half as likely as married fathers to be in full-time employment (53 per cent and 87 per cent, respectively; χ2 = 12.114; d.f. = 3; p < 0.01).4 While there were no differences in employment status between cohabiting and married mothers, cohabiting mothers were twice as likely as married mothers to be in the working class (67 per cent and 33 per cent, respectively). These findings are consistent with previous studies (McRae 1993, 1999; Smart and Stevens, 2000; Lewis, Papacosta and Warin, 2002; Ermisch, 2005), showing that cohabiting parents may therefore replace or mirror some of the unstable shotgun marriages of former times, as well as with Holtermann et al’s (1999) analysis of the 1997 Labour Force Survey, 4 The 1994 data indicate that our 2002 findings are likely to underestimate the differences in employment status and income between cohabiting and married parents, because a few post-marital cohabitants may have been included in the category of cohabiting parents. Post-marital cohabitants, however, are far more likely than cohabiting parents to both be in full-time employment. In 1994, for example, differences in income between married and cohabiting parents were much bigger (£15,406 and £7,348 per annum, respectively; p = £100 pa High education Working Hours worked Holds second job Marriage tenure Previously married Cohabited before marriage Characteristics at t = 5 Partner New child Started work Stopped work Retraining (if no children) Retraining (if child present at t = −1) Change in hours Mean probability of repartnering Observations

Growth in Household Income −0.030 −0.154 0.449 −0.398 0.079 −0.090 0.064 0.102 −0.212 0.008 0.446 0.026 −0.226 0.273

−0.019 0.126 0.581 −0.221 0.088 −0.153 0.081 −0.027 −0.299 0.021 0.583 0.022 −0.252 0.163

Probability of Repartnering −0.028 −0.219

−0.026 0.002

−0.088 −0.019 0.037 0.188 −0.026 0.040 −0.005 0.159 0.022 0.027 0.451

−0.189 −0.015 0.160 0.101 −0.027 −0.075 −0.008 0.128 0.016 −0.080 0.431

0.409 −0.094 −0.068 −0.017 0.138 −0.534 0.021

−0.399 −0.195 −0.012 0.021 −0.171 −0.004 0.430

204

0.424 200

Notes: For columns 1 and 2, the coefficients on each variable show the effect on the growth of income from the year prior to separation to five years after separation. For example, the coefficient on the variable ‘Cohabited before marriage’ indicates that the rate of income recovery is 0.27 higher (ie 27% higher) five years after divorce for those who cohabited prior to marriage compared with those that did not, after controlling for other characteristics. Column 1 includes as explanatory variables only those variables which were known at the time of divorce, column 2 adds as explanatory variables characteristics of the individual five years after divorce (these characteristics will be affected by what happens to income and so their interpretation is less clear cut). For columns 3 and 4, the coefficients on each variable show the change in the probability of repartnering for a one unit change in that variable, relative to the probability of repartnering for an individual with average characteristics (and the average probability of repartnering). The coefficient ‘Cohabited before marriage’, for example, shows that those who cohabited before marriage have a probability of repartnering after divorce which is 0.451 (ie 45%) higher than those who did not cohabit before marriage. Variables in bold are significant at a 10% confidence level. Variables which are not in bold are statistically insignificantly different from zero. Other controls: changes in home ownership and health status. Further details and standard errors available on request.

244 Hayley Fisher and Hamish Low This analysis suggests that those groups of women who suffer most from divorce are those who are older and those who have children. Part of the reason for their inability to recover financially compared with others is the lower chance of repartnering. The persistence of the fall in equivalised household income for these women highlights the importance of divorce in driving outcomes for women and their children. Changes in Effects Over Time Our data covers separations occurring over a 15-year period. However, our analysis so far has assumed that the effects of divorce are constant over time and that the average income fall on separation is the same in the early 1990s as in 2004. Over this period there have been various social changes including increasing divorce rates (albeit at a slower rate than previously), falling marriage rates, and increased female labour force participation, as well as some key House of Lords rulings regarding the division of assets at the time of divorce,18 and we might therefore expect that the extent of the income fall on divorce has changed over time. Due to the limited sample size we are unable to run the regressions using a full set of time-to-divorce and year dummy interactions. Instead, we run the regression with equivalised household income as the dependent variable on two sub-samples: an earlier period covering separations from 1992 to 1997, and a later period covering separations from 1998 onwards. The results for women are presented in Table 11.4. We see clearly that the impact of divorce is reduced in the later period, with women suffering a 42 per cent fall in income relative to a 65 per cent fall in the earlier period. Recovery is, however, less pronounced in the later period. This suggests improved insurance against divorce in the period after 1997 as compared to the 1990s, which might be attributed to higher labour force attachment. This is consistent with McKeever and Wolfinger’s (2001) findings for the US, and also with Jenkins’ (2008) findings for the UK, which stress the increased role of in-work benefits as a means of providing insurance. For women, the positive effect of repartnering remains, as does the negative effect of having a child. Unreported results for men, however, show that their experience on divorce does not change over the two periods. This analysis is suggestive of a reduction in the divergence between the experience of men and of women after divorce: women have become better insured against the income loss, and men do not make the gain in equivalised income they did previously. FINANCIAL LOSS FOLLOWING COHABITATION

Due to sample size restrictions, a number of previous studies have estimated income changes for a sample combining individuals who separated from marriage 18 White v White [2001] AC 596, and Miller v Miller; McFarlane v McFarlane [2006] UKHL 24, [2006] 2 AC 618.

Who Wins, Who Loses and Who Recovers from Divorce? 245 Table 11.4: Explaining Income Growth for Women in Different Time Periods To 1997 t = −1 t=0 t=1 t=2 t=3 t>3 Repartnered Partner Child present Home owner Asset income > £100 pa Working Receives transfer Makes transfer Observations People

From 1998

−0.026 −0.648 0.146 −0.048 −0.004 −0.048 0.430 0.001 −0.090 −0.005 −0.090 0.071 0.031 −0.009

0.016 −0.417 0.069 −0.036 −0.005 0.040 0.456 0.023 −0.075 −0.042 −0.051 0.047 0.035 0.125

2,368 213

1,398 204

Note: The coefficients on each variable show the effect on the rate of growth of income of that variable. For example, the coefficient on the variable t = 0 in the first column indicates that the rate of income growth is 0.65 lower over the time of separation (ie income falls by 65% on separation) in the period before 1997, but the coefficient in the second column indicates only a 42% fall for those who divorce after 1997. Bold denotes variables which are significant at 10% confidence level. Variables which are not in bold are statistically insignificantly different from zero. Other controls: age at separation and age at marriage, current age, marriage tenure, previous marriage, education, race, prior cohabitation, poor health and holding a second job. Further details and standard errors are available on request.

and those who separated from cohabitations (see, for example, Jenkins, 2008). Our larger sample size allows us to make comparisons between income changes resulting from splits from marriage and cohabitation. Figure 11.6 shows the evolution of log equivalised household income for men and women, split into those separating from cohabitations and those from marriage. The graph shows the change in log income since time t = –1. On average, the income fall is smaller for those splitting from cohabitations. This might be explained by the fact that only 36 per cent of women separating from cohabitations have children, compared with 69 per cent of those from marriage. Further, those splitting from cohabitation tend to be younger than those divorcing. Recovery from cohabitation also occurs more quickly. We do not see the same instant jump upwards in equivalised household income for men splitting from cohabitations as described above for men divorcing their wives. This is consistent with the lower prevalence of children, meaning that

-.6

Log Income (rebased) -.4 -.2 0

.2

246 Hayley Fisher and Hamish Low

-1

0

1

2 3 time to separation

4

5

Male Income (married at t=-1)

Male Income (cohabiting at t=-1)

Female Income (married at t=-1)

Female Income (cohabiting at t=-1)

Note: The four lines show the average of (log) income by number of years since separation for men and for women, separately for splits from marriage and for splits from cohabitation, controlling for changes in household size (equivalised income). Income in each series is rescaled to begin at 0 at time t = –1. The scale on the y-axis should be read as the proportional change in income relative to time t = –1. For example, for women, the proportional fall in equivalised income at the end of cohabitation is 0.2 (ie a 20% decline) compared with a 50% decline at the end of marriage. Figure 11.6: Evolution of Income after Split from Marriage vs Cohabitation

the equivalence scale changes less. However, the differences between cohabitation and marriage persist even when we control for age, duration and children as in Table 11.2: comparing the income losses for cohabitants with those for divorcees as in Table 11.2 shows an average loss on separation of 32 per cent in equivalised income for cohabitants compared with 57 per cent for divorcees, and a 45 per cent loss in household income for cohabitants compared with 83 per cent for divorcees. These figures indicate that a split from cohabitation is associated with a smaller economic impact than a split from marriage. This is especially pronounced for women, who experience much smaller falls in income, and recover more quickly than from divorce. It is difficult to draw stronger conclusions about the evolution of income over time after the separation because the data become more noisy as time since separation increases. However, our data suggests that a split from cohabitation is a less traumatic experience in terms of household income than a divorce.

Who Wins, Who Loses and Who Recovers from Divorce? 247 This is consistent with the analysis of Manting and Bouman (2006), who studied data from the Netherlands from 1989 to 2000, finding that cohabiting women experience a smaller income decline on cohabitation, and also that gender differences after the split disappear over time when considering cohabitation. Avellar and Smock (2005) find a similar pattern in the US. Even though our conclusion about the difference between divorce and the end of cohabitation persists when we control for observable characteristics, it is important to stress that there may be differences between the types of person who marry and those who cohabit which may underlie this conclusion: a woman divorcing from marriage may in fact be better off than she would have been had she been splitting from cohabitation. This would arise if the population of those splitting from cohabitations are less vulnerable (in terms of income) to separation than the population of those divorcing. This would imply that had those divorcing instead separated from a cohabitation, they would have seen an even larger income fall owing to the lack of financial remedy from the male partner. THE EFFECTS OF DIVORCE ON INCOME AND CONSUMPTION

The conclusion from our analysis of the BHPS is that recovery from separation occurs, but that it can be slow, and is less likely for older women or when children are present. This motivates the discussion of possible bases for settlement. There are therefore three aims of this section: first, to show the implications for earnings of alternative decisions within marriage; secondly, to show the implications for income post-divorce of alternative mechanisms for allocating resources; and, finally, to discuss the implications of the alternative allocation mechanisms for consumption. In discussing alternative methods for calculating divorce settlements, the focus is on methods for allocating earned income and assets combined. There is discussion in the divorce literature of the desirability of ‘clean breaks’. There are two separate issues: one is the desirability of calculating the amount of the settlement at the time of the divorce; the second is the desirability of making all payments associated with a settlement at the time of divorce. Whilst full payment clean breaks would have the advantage of allowing the individuals to go their separate ways and prevent any dependency on the ex-spouse, such an arrangement is not feasible when human capital is an important component of wealth. Wages and the Return to Work To make the analysis more transparent, we make the following assumptions, although the conceptual framework does not require these assumptions: A1 The potential income process faced by husband and wife is identical at the time of marriage, that is, they have the same career prospects and so

248 Hayley Fisher and Hamish Low it may be supposed that the wife could have had the same future income as the husband; A2 No individual assets prior to marriage and no inheritances; A3 No retirement. Figure 11.7 shows potential earnings at each age following different decisions about participation in the labour force. We consider what happens to potential wages after exit from the labour force. Individuals who remain single and continue working full time are estimated to increase their wages each year by about 2.4 per cent in real terms (Eckstein and Wolpin, 1989). After 10 years of work, at D, their wages will have increased by 27 per cent. For married individuals, there is potentially an extra return, leading them to E. Antonovics and Town (2004) estimate the benefit to marriage as a 20 per cent wage premium, after controlling for selection, and show that this premium persists after divorce. Other evidence from the US (Korenman and Neumark, 1991) suggests that only a small portion of this benefit derives from support from a wife who is not working, with the rest accruing from marriage whether or not the spouse works. This is therefore a benefit of being married independent of the decision to work or not. However, this finding is not uncontroversial. For example, Bardasi and Taylor (2005) analyse the male marriage premium for the UK using BHPS data, using a fixed effects approach, and find a much smaller marriage premium (around four per cent), which declines with the hours of market work undertaken by the wife. The source of the marriage premium is important to determine whether there is double counting of the effects of decisions within marriage: double counting would arise if both the loss to the wife from not working and the ongoing gain to the husband following divorce arising from the support provided by the wife during the marriage are

Earnings

Full-time and supported E D A

Full-time Lost opportunity only Lost opportunity and depreciation

B

30 Exit

40 Reentry

Figure 11.7: Earnings over the Life-cycle

Age

Who Wins, Who Loses and Who Recovers from Divorce? 249 compensated (in the latter case, by sharing that gain). Double counting would not arise if the gain from the wife’s support was simply a marriage premium which was independent of whether the woman was working. Those who exit from the labour force are not able to obtain the wage increases from extra experience, and, in addition, their skills are likely to depreciate. Without depreciation, they would be at A after 10 years, with potential earnings 66 per cent of those of their partner. With a depreciation rate of two per cent a year (as estimated in Mincer and Ofek, 1982), potential earnings on returning to work after 10 years are 54 per cent of their partner’s earnings.19 These numbers are, of course, only indicative and are based on average returns to experience and depreciation and ignore any gender wage gap. In reality, returns to experience are likely to be age dependent and higher for younger workers. This would make the earnings deficit harder to recover from for individuals who are older when re-entering the labour market. The numbers also differ substantially across occupation and education groups: better-educated workers tend to receive greater returns to experience and are subject to faster wage depreciation.

Divorce Payments This discussion of the cost of non-participation is in terms of lost realised earnings and lost earnings potential. Clearly, however, there are benefits of not participating in the labour market, both from the support this may offer to a partner, and from not having to pay for child care. To the extent that costs and benefits are both realised within the marriage, divorce payments should be independent of such costs and benefits. The difficulty is that decisions within marriage have costs and benefits after divorce. Figure 11.8 puts Figure 11.7 into the context of divorce. In this example, divorce happens before re-entry into the labour force. At the time of divorce, the earnings of the partner who has kept working are at E, whereas the wife, who exited 10 years previously, is at 54 per cent of this level, at B. This difference can be broken down into the higher wage that arises if married (z1) and the opportunity cost of not working (z2). The amount z1 is referred to as the ‘restitution damages’ and the amount z2 as ‘reliance damages’ (Dnes, 1999). However, using z1 and z2 as calculations of the benefit and cost of not working underestimate both the cost and the benefit. The boost to wages from being supported within marriage persists after divorce, generating the benefit x1 and y1. Similarly, the cost to wages from not working persists beyond divorce, with the additional cost to potential earnings of x2 + y2. However, actual earnings after divorce depend on when the non-working partner re-enters work, with each year of non-participation lowering potential wages further. In the figure,

19 If we ignore the ‘marriage premium’ benefit accrued by the working husband where the estimates are less reliable, potential earnings would be 65% of the husband’s.

250 Hayley Fisher and Hamish Low Earnings

Full-time and supported x1

Full-time

y1

x2

Potential Earnings

y2

x3

Actual Earnings

E z1 z2 y3 B y4

30 Exit

40

x4

Age

43 Reentry

Divorce

Figure 11.8: Divorce and Earnings over the Life-cycle

this occurs three years after divorce, when the potential wage has fallen a further 6 per cent.20 These decisions mean that total resources post-divorce are given by: Total =

x4


Wife’s actual income

+ x4 + x3 + x2 + x1 + y 4 + y 3 + y 2 + y1  Husband’s actual income

The realised monetary loss post-divorce as a result of non-participation (that is, the difference between actual earnings after divorce and potential earnings if there had been no period of exit) is given by: Loss = x3 + x2 + y 4 + y 3 + y 2 However, this is misleading as an estimate of the cost of decisions within marriage, because part of this loss is due to decisions after divorce. The monetary loss we should be considering is the impact on potential earnings of decisions within marriage. This is given by: Potential Loss = x2 + y 2

20 We can relax the assumption of equal income processes, so that, eg, the spouse with the greater earning potential works while the other takes time out of the labour market. The simplest way to model this is to assume that Figure 11.8 shows the earnings process for the spouse who is not working, whereas the spouse who is working has a greater income. This implies that the total resources postdivorce will be greater than if earnings are equal. However, the potential earnings post-divorce of the non-working spouse will not vary with the income of the working spouse.

Who Wins, Who Loses and Who Recovers from Divorce? 251 This potential loss is the ‘compensable loss’ to the woman. An allocation mechanism based on compensation would split this loss 50:50 between the husband and wife. In the example where: (i) the marriage lasted 10 years; (ii) the wife was not working for any of that period; and (iii) potential earnings before marriage were identical, then the husband would be paying approximately 17 per cent of his earnings.21 This, however, ignores the ongoing benefit to the husband of marriage. Allowing for these benefits generates the difference in potential earnings that arises from the marriage: Difference in potential =

x2 + y 2    

Loss to wife’s potential earnings

+

x1 + y1  

Increase in husband’s potential earnings

This would imply the husband making a payment of 23 per cent of earnings in the example above. If there are benefits that accrue to the husband only if his wife is not working, then either the costs or the benefits should be used in determining the amount, but not both. To focus on the incentive effects of this allocation, we start by assuming that the decisions to marry and to divorce are unrelated to current earnings and earnings potential, and that decisions about labour market participation do not affect the bargaining position within marriage. The allocation mechanism then has a potential incentive effect on labour market participation and on spousal support within marriage. A mechanism which under-rewards the benefit of marriage to the husband’s career that persists after divorce (underestimating x1 + y1) or underestimates the loss after divorce to potential earnings of not working within marriage (underestimating x2 + y2) will impose financial losses disproportionately on the woman. This is the context in which the McFarlane ruling should be assessed. The McFarlane22 ruling can be interpreted as arguing in favour of using estimates of this ‘difference in potential’: the marriage had been long and Mrs McFarlane had stopped work to bring up the children, thereby incurring the loss of earnings potential, in addition to the support provided to her husband. If this difference in potential earnings were not accounted for in divorce settlements, then the lost earnings potential could be mitigated by increasing labour market participation within marriage. If this were the case, then the choice within marriage would be distorted towards excess labour market participation as a form of protection or insurance against the possibility of divorce. By contrast, if the compensable loss were correctly calculated, there would be no distortion to the labour market participation decision. In

21 This assumes that if the wife had worked, her wage would have grown by 3% per year, instead of declining by 2% per year. After 10 years, the obtainable wage is 60% of the wage that would have been earned and this difference is then split 50:50. 22 [2006] UKHL 24.

252 Hayley Fisher and Hamish Low this sense, the McFarlane ruling reflects the removal of a distortion to decisions within marriage.23 Two other points about the McFarlane ruling are important in this context: first, there is a focus on the retraining possibilities after divorce and a suggestion that the payments should be time limited. This would be appropriate from an economic perspective if retraining could offset the lost opportunities within marriage and if the higher wage for men associated with being married did not persist beyond divorce. However, the evidence is that retraining is only partially effective and that the higher wages associated with marriage are highly persistent after divorce. Secondly, there was concern in the ruling about payments being used for savings, but since savings are a mechanism for deferring consumption, the use of a payment would not be relevant if earnings were the basis for the calculation. It is, however, worth distinguishing between basing compensable loss on potential earnings after divorce compared with actual earnings after divorce. If the basis for calculating loss were actual earnings, then the choice by the woman not to work would impose a cost on her ex-husband for a decision which was not a joint-decision. By contrast, potential earnings after divorce include the implications of decisions to work within marriage, but leave the choice and consequences of whether or not to work after divorce solely with the woman. Dnes (1999) argues that one of the principal difficulties with using allocation mechanisms based on the costs and benefits of decisions within marriage is that it may make divorce too cheap for one party relative to expectations of the cost of continuing with the marriage. This arises in particular when the underlying earnings potential of the husband and wife differ. This would give rise to opportunistic behaviour and the incentive for the higher-earning partner to divorce and thereby renege on the implicit agreement within the marriage. Discussion of such implicit agreements and expectations raises the issue of allocation on the basis of consumption rather than earnings loss. Allocations on the Basis of Consumption An alternative to making divorce payments dependent on earnings or earnings loss is to use a measure of consumption (or spending power). The rationale for this lies in a desire to use allocation mechanisms to ensure needs are met, or insurance is provided against divorce, or that allocations should be set to compensate for lost expectations. However, this is the issue of whether divorce allocations are about maintaining a standard of living and honouring a commitment within marriage, as compared to ensuring an appropriate allocation of the benefits and costs arising from the marriage. The former leads to a consumption-based rule, the latter to an income-related rule. 23 One criticism of the equal split of the costs and benefits of decisions within marriage is that the incentive to divorce lies with the party who in practice bears more than 50% of the net costs.

Who Wins, Who Loses and Who Recovers from Divorce? 253 Irrespective of the mechanism, it is important to understand the effects of divorce on consumption because this is more closely related to welfare than income. The main point about divorce in this context is that even if resources are split equally, the standard of living of each individual will fall because it is cheaper for two people to live together than separately: divorce means that these economies of scale from sharing a house are lost. This was highlighted on pp 235–238 above, where allowing for household size and composition was shown to be crucial to conclusions about the consequences of divorce. This is reiterated in Figure 11.9, which superimposes consumption choices pre- and post-divorce. To make this simple, we assume that consumption (or spending power) does not vary over time before divorce and does not vary over time after divorce (although consumption after divorce may be at a different level). On divorce, if we ignore saving, the total consumption of the former couple will fall because of the lost economies of scale, and this is reflected in the fall to cDiv,Equal in the figure. This is the heart of the problem for divorce: there is insufficient income for both parties to maintain their pre-divorce standard of living. It is, of course, possible for one party to maintain their standard of living if the other party receives less, with the lowest level of spending power being the consumption level from complete independence (or autarky), labelled on the figure as cDiv,None. Indeed, as shown earlier, men do better on these measures after divorce than they did within marriage. A further issue is that these graphs showing potential earnings and actual earnings ignore the possible presence of fixed costs of work associated with child care. This suggests that the graphs overstate the potential earnings (and consumption) of the spouse, and the earnings net of child-care costs are likely to be substantially lower. An alternative way of allowing for child-care costs associated with work would be to calculate the utility value of alternative income paths. This would earnings, consumption

Full-time and supported

Potential Earnings Actual Earnings

cMarried cDiv, Equal

cDiv, None

Exit

Reentry Divorce

Figure 11.9: Consumption and Earnings over the Life-cycle

age

254 Hayley Fisher and Hamish Low recognise explicitly that the value of different income streams may differ between spouses depending on their obligations. In principle, this notion could be made more precise by considering consumption paths over the life-cycle and the impact of divorce on those consumption paths. Implicitly this recognises that the costs of decisions taken within marriage (for example, having children) should be split between the parties—exactly analogously to the argument for splitting the compensable loss to earnings. Where there are substantial differences in the underlying earnings potential or asset holdings of the two parties, the compensation basis for settlement makes divorce ‘cheap’ for the wealthier party, as discussed above. Therefore, this discussion of allocation on the basis of spending power is most appropriate when there are substantial differences in the underlying earnings potential or asset holdings of the two parties, as with the Miller24 ruling. The Miller case arose from a childless marriage of short duration. The consequences of decisions within the marriage for earnings would appear to be minimal in this case, arising only from the increase in Mr Miller’s earnings during the marriage. It is hard not to draw the conclusion that the basis for settlement was maintaining the consumption level for Mrs Miller that had implicitly been promised by Mr Miller. This would imply that the judgment was based on an assessment of what the implicit contract of marriage is, rather than being based on the economic consequences of decisions within marriage. However, the notion of legitimate expectation is explicitly excluded and so the basis of the ruling is less clear. Further, although the settlement was substantial, Mrs Miller would not have been able to maintain her level of consumption over the longer term. The incentive problem with a legitimate expectation ruling is that marriage becomes potentially very costly when there are differences in earnings potential—the basis for divorce settlement will act to discourage marriage in these cases. CONCLUSIONS

The aim of this chapter was to analyse the economic consequences of divorce. The stark conclusion is that men’s household income increases by about 23 per cent on divorce once we control for household size, whereas women’s household income falls by about 31 per cent. There is partial recovery for women, but this recovery is driven by repartnering: the average effect of repartnering is to restore income to pre-divorce levels after nine years. Those who do not repartner tend to be older and have children. For these individuals, and for those in poor health at the time of divorce, the long-term economic consequences of divorce are serious. For some, these long-term consequences are offset by increased labour supply, but the effects are small, and this ignores any extra costs such as child care which

24

[2006] UKHL 24.

Who Wins, Who Loses and Who Recovers from Divorce? 255 may arise from working. For others, government-provided benefits provide some cushion to the cost of divorce. On the positive side, we present evidence that these costs of divorce for women have been mitigated over time and more recent divorces have not led to the same falls in household income as earlier divorces. These conclusions on the financial consequences of divorce highlight the importance of the basis on which post-divorce settlements are calculated. The McFarlane case is an example of the scenario when the economic consequences of divorce are most severe for the woman. The key change highlighted by the McFarlane ruling was an attempt to consider the impact on post-divorce income arising from decisions within the marriage. In particular, this involved considering the lost return to work experience that would have occurred if Mrs McFarlane had kept working and considering the increase in her husband’s post-divorce earnings due to the marriage. Correctly calculating the difference to post-divorce earnings arising from decisions within marriage removes some of the distortions to decisions about labour supply within marriage. However, given the difficulty of these calculations and given the difficulty of accurately modelling behaviour within relationships, these are problems that are not going to go away.

BIBLIOGRAPHY Aassve, A, Betti, G, Mazzuco, S and Mencarini, L (2006) ‘Marital Disruption and Economic Well-being: a comparative analysis’, ISER Working Paper 2006–07 (Colchester, University of Essex). Antonovics, K and Town, R (2004) ‘Are All the Good Men Married? Uncovering the Sources of the Marital Wage Premium’ 94 American Economic Review 317. Avellar, S and Smock, PM (2005) ‘The Economic Consequences of the Dissolution of Cohabiting Unions’ 67 Journal of Marriage and the Family 315. Axinn, WG and Thornton, A (1992) ‘The Relationship between Cohabitation and Divorce: Selectivity or Causal Influence?’ 29 Demography 357. Bardasi, E and Taylor, MP (2005) ‘Marriage and Wages: A Test of the Specialisation Hypothesis’, ISER Working Paper 2005-01 (Colchester, University of Essex). Chiappori, P-A, Fortin, B and Lacroix, G (2002) ‘Marriage Market, Divorce Legislation, and Household Labor Supply’ 110 Journal of Political Economy 37. Dnes, AW (1999) ‘Applications of economic analysis to marital law: concerning a proposal to reform the discretionary approach to the division of marital assets in England and Wales’ 19 International Review of Law and Economics 533. Duncan, GJ and Hoffman, SD (1985) ‘Economic Consequences of Marital Instability’ in M David and T Smeeding (eds), Horizontal Equity, Uncertainty, and Economic WellBeing, Studies in Income and Wealth vol 50 (Chicago,IL, National Bureau of Economic Research). Eckstein, Z and Wolpin, K (1989) ‘Dynamic Labour Force Participation of Married Women and Endogenous Work Experience’ 56 Review of Economic Studies 375. Eekelaar, J and Maclean, M (1986) Maintenance after Divorce (Oxford, Clarendon Press). Jarvis, S and Jenkins, SP (1999) ‘Marital splits and income changes: Evidence from the British Household Panel Survey’ 53 Population Studies 237.

256 Hayley Fisher and Hamish Low Jenkins, SP (2008) ‘Marital splits and income changes over the longer term’, ISER Working Paper 2008-07 (Colchester, University of Essex). Korenman, S and Neumark, D (1991) ‘Does Marriage Really Make Men More Productive?’ 26 The Journal of Human Resources 282. Lillard, LA, Brien, MJ and Waite, LJ (1995) ‘Premarital Cohabitation and Subsequent Marital Dissolution: A Matter of Self-Selection?’ 32 Demography 437. Manting, D and Bouman, AM (2006) ‘Short- and Long-term Economic Consequences of the Dissolution of Marital and Consensual Unions. The Example of the Netherlands’ 22 European Sociological Review 413. McKeever, M and Wolfinger, NH (2001) ‘Reexamining the Economic Costs of Marital Disruption for Women’ 82 Social Science Quarterly 202. McManus, PA and DiPrete, TA (2001) ‘Losers and Winners: The Financial Consequences of Separation and Divorce for Men’ 66 American Sociological Review 246. Mincer, J and Ofek, H (1982) ‘Interrupted Work Careers: Depreciation and Restoration of Human Capital’ 17 The Journal of Human Resources 3. Smock, PJ, Manning, WD and Gupta, S (1999) ‘The Effect of Marriage and Divorce on Women’s Economic Well-Being’ 64 American Sociological Review 794. Taylor, M, Brice, J, Buck, N and Prentice-Lane, E (2007) British Household Panel Survey User Manual Volume A: Introduction (Colchester, University of Essex). Uunk, W (2004) ‘The Economic Consequences of Divorce for Women in the European Union: The Impact of Welfare State Arrangements’ 20 European Journal of Population 251.

12 Pension Accumulation and Gendered Household Structures What are the Implications of Changes in Family Formation for Future Financial Inequality? DEBORA PRICE

INTRODUCTION

A

s the uk population has rapidly aged over the last decades, marked gender disparities in income and poverty in later life have been exposed. Older women have much less income and far higher rates of poverty than men (Arber and Ginn, 2004; DWP, 2005). These inequalities are the result both of gendered differences in the life-course and a pension system designed to reward full-time employees with full working histories working for large employers. Women have historically been much more likely than men to undertake care work and housework within the household, to work part time, for low pay, for small employers, and to have interrupted histories of paid work. They have been correspondingly less likely to participate in state, occupational and private pensions, or to remain in the paid workforce as they approach state pension age. Particularly noticeable, however, is the low income of married older women (Ginn, 2003). It is this low income within marriage that accounts ultimately for the relatively high rates of poverty among older female divorcees and widows, both disproportionately poor groups in later life, as they cannot compensate for the loss of their financial dependency on their former husbands (Price, 2006). This is so even though widows usually inherit some pension rights from their deceased husbands, unlike divorcees, who are the poorest group of older women (Arber and Ginn, 2004). The question arises whether this will remain so for future cohorts. Many aspects of social life have changed dramatically, particularly the increasing education, labour force participation, earnings and independence of women. As rates of cohabitation and divorce have also risen over the decades, this too poses questions for understanding the cultures and structures that might affect older women’s

258 Debora Price income in later life. Legal marriage provided some protection in terms of derived social security, pension, and widow’s benefits, but partnering outside legal marriage has grown, those living alone have increased in number, and the incidence of separation and divorce has risen. If women are potentially acquiring independent incomes in later life through their own pension accumulation regardless of their marital status, then we need not perhaps be too concerned about changes in family formation and dissolution for the later life income of women. If, however, women in couples continue to exhibit high levels of financial dependency on their male partners within marriage or cohabitation, then the consequences for later life income of all women bear further examination. Behaviour within couples affects not only the individuals concerned, but (in circular feedback loops to policy and from policy back to behaviour) will also have an impact on the policy environment, the labour market and gendered hours of working, child-care provision, and patterns of partnering and dissolution (Lewis, 1992, 1997; O’Connor, Orloff and Shaver, 1999; Pfau-Effinger, 1999; Orloff, 2002). These wider social and policy impacts will also affect the ability of women who are not in couples to earn and to accumulate pensions. Moreover, cultures of gendered financial dependency within couples will probably themselves impact on long-term financial decision-making. Qualitative research has shown that women who are dependent on men for income are generally also dependent on them for pensions, despite generally high risks of the partnership breaking down (DWP, 2005). In a study of women’s pension choices in the mid 1990s, Peggs (2000: 356) found that many women saw their employment as temporary and ‘second to caring’. Without taking on a breadwinning role, which involved planning ahead financially, these women did not—and with low earnings had few options to—participate in private pension schemes. In a comprehensive study of women’s pensions in Northern Ireland, Evason and Spence (2002: 35–41) found that older cohorts of women depend heavily on their husbands for pension provision; similarly, while younger cohorts would have liked to have been independent of their partners in providing pensions for themselves in retirement, or even felt that they ought to be independent, where they had children this was simply not the case, and they thus continue to rely on their husbands or partners to accumulate pensions for both of them. Understanding cultural patterns of earnings and pension accumulation within working-age couples is therefore important for understanding potential gender disparities in later life income. Yet despite its importance, very little research has examined these issues. In this chapter, I use data from the General Household Surveys (GHS) 2001 and 2002 to address five questions: (1) To what extent do couples in the UK still organise themselves around breadwinner models of coupledom? (2) What is driving these modes of organisation? In particular, does the degree of financial inequality between partners vary according to the woman’s level of educational attainment?

Pension Accumulation and Gendered Household Structures 259 (3) Is the extent of financial inequality within couple relationships related to the extent to which women participate in pension accumulation? If so, what are the drivers for this? (4) How, if at all, is financial inequality within relationships related to legal marital status and history? (5) What are the implications of these modes of organisation for the pension accumulation of working-age women, whether single or in a couple? These questions are all asked within the context of a social structure in the UK where women are particularly likely to work part time for low pay, in a system for pension accumulation which privileges those in full-time work on high pay.

COUPLES, EARNINGS, AND NORMS OF PARTNERING

In a low-waged economy for women coupled with a lack of state support for care and caring roles in the UK, household decisions are often made for women to prioritise within-household family care over paid work (see further Scott and Dex, chapter three, this volume). While between 60 and 70 per cent of ‘working-age’ women in the UK are in the paid labour force, 45 per cent of these work part time (Women and Equality Unit, 2005). This pattern has been remarkably resilient to change over time (Woods et al, 2003). Research also confirms that part-time work for women in the UK is often not directly associated with child care (Johnson and Stears, 1996; Walby and Olsen, 2002; Manning and Petrongolo, 2004: table 1 and figure 3.2); almost 30 per cent of women with no child under 16 in the household work part time, compared with 8 per cent of men (Women and Equality Unit, 2005). Women may well be making the best of what is available to them, given the competing demands on their lives and on their time (Warren, 2004). These patterns nevertheless point to a model of gender relations within families that is socially, normatively, and structurally embedded. Policies supporting parttime work often reinforce the identity of women as carers (McKie, Bowlby and Gregory, 2001), and serve to reinforce gender differences in pay, prospects, and conditions (Smithson et al, 2004). The extent of part-time working for women in the UK suggests that women probably remain financially dependent on men within couples to some extent (see further Vogler, chapter four, Burgoyne and Sonnenberg, chapter five, and Lewis, Tennant and Taylor, chapter eight, this volume). Women’s dependency arises from the extent to which they are unequal contributors to the household (Ward, Dale and Joshi, 1996b). They are not necessarily ‘dependent’ in the sense of being unable to form an autonomous household (although these two forms of dependency will often coincide), but they have much to lose from the ending of the relationship (Millar, 2003). The lifestyle of many women is closely tied to the financial resources provided or sustained by their partners, including housing, furniture, household goods, cars, holidays, entertainment, and other material things.

260 Debora Price Warren (2001: 555) describes women in this position as having a ‘distinctly contingent role’. Full-time work is recognised as the route by which women generally achieve greater financial independence from their partners (Ward, Dale and Joshi, 1996a), although even among full-time workers, disparities in working hours and in pay may leave couples with wide disparities in earnings. Using data from the late 1980s, Arber and Ginn (1995) showed that gender inequality in earnings was more pronounced within households than in society generally. They found that among wives aged 20 to 59, about a third were not in paid work, with just over a third working part time, and the remainder working full time. Among dual-earning couples, a quarter were approximately equal earners, and only 11 per cent of wives earned more than their husbands. Similar findings were obtained in an analysis of couples in the 1958 cohort at age 33 in 1991 (Ward, Dale and Joshi, 1996b), and then again by Warren (2003) using 1995 data from the British Household Panel Study. Warren showed that among working-age dual-earning couples a third of men earned more than 75 per cent of household wages, just over half earned more than 65 per cent, and three quarters more than 55 per cent. Only 9 per cent of women in dual-earning couples earned more than 55 per cent of combined wages. These three studies paint a picture of substantial and surprisingly static continuing inequalities of earnings within couples. Given these patterns, what are the implications of this inequality for women’s participation in private and occupational pension schemes?

FINANCIAL DEPENDENCE IN LATER LIFE: A GENDERED PENSION REGIME

The UK pension system is a mixture of insurance-based state provision (through the system of National Insurance, and National Insurance credits), other state benefits (including means-tested benefits), and occupational or private pensions. The Pensions Commission has described the system as one of ‘bewildering complexity’, indeed, as the most complex pension system in the world (Pensions Commission, 2004: xiii, 238). The system has been designed over many decades to encourage and incentivise private and occupational pension provision over state provision, so that citizens are largely dependent on private and occupational pensions for adequate income in later life (Glennerster, 2006). The state pensions (basic state pension and state second pension) together provide an income well below the poverty line for most people, but particularly for women (Arber and Ginn, 2004; Price, 2006). This means that citizens in the UK are particularly reliant on employer or personal private provision for adequate income in old age. Men over 65 have far higher occupational and private pension incomes than women (Arber and Ginn, 2004; Ginn, 2007) exacerbating gender differences in income in later life. Women are therefore generally reliant on a partner’s pension, or, if they live alone, on means-tested state benefits, for an adequate income in retirement. Means-testing

Pension Accumulation and Gendered Household Structures 261 is, however, a flawed system for reaching those most in need. Somewhere between 30 per cent and 40 per cent of pensioners who are entitled to Pension Credit (the principal means-tested benefit for pensioners) do not claim their entitlements (DWP, 2007). The government has spent the last five years shaping reform of the pension system. The reformed system will have an improved basic state pension, but this will still provide income well below the poverty line, and participation in paid work throughout the life-course, and in private or occupational pensions, will remain central to building adequate income in later life (Price, 2007, 2008). Analysis in this chapter is therefore of participation in ‘third-tier’ pension provision, meaning occupational and private pensions over and above the current (low) levels of mandatory provision in the state pension schemes or their compulsory equivalent. It is participation in this tier of pension scheme provision that will determine the extent to which people have an adequate income for a comfortable later life. DATA AND METHODS

The analysis combines cross-sectional data from two years of the General Household Surveys (2001 and 2002). The General Household Survey is a multipurpose continuous cross-sectional survey carried out by the Office for National Statistics, collecting information on a range of topics from people living in private households in Great Britain, including current pension scheme participation, marital and cohabitation histories, and maternal histories. A stratified, clustered probability sample of about 13,000 households is selected, and each adult individual in the household is interviewed. The survey achieves response rates of approximately 70 per cent. The combined data set yields 10,314 men and 11,087 women aged between 20 and 59. Included in the data is information about the earnings of 5,772 partnered men and 6,141 partnered women aged 20 to 59.1 Where contributions to private pensions are shown in this chapter, these are either occupational pension schemes provided by an employer, or personal pensions provided by an insurer, but if the personal pension is only a contracted-out replacement for the state scheme, then it has been excluded. Those included are described here as ‘third-tier pensions’. Women have been grouped into those who have never had children, those whose youngest child is under six, those whose youngest is between 6 and 16, and those whose youngest is over 16. For individual earnings analysis, all those in paid work have been divided into quintiles according to gross weekly earnings. Partnerships have been classified according to the combined earnings of partners. Those partnerships in receipt of any income from earnings have again been divided into joint earnings quintiles.2 1 2

Same-sex partnerships are excluded from the analysis. The cut points are available from the author.

262 Debora Price EARNINGS INEQUALITIES WITHIN COUPLES: THE MODERN BREADWINNER

Men

.74 .72 .70 .68 .66 .64 .62 .60 20-24

25-29

30-34

35-39 40-44 Age Group

45-49'

50-54

55-59

Mean Ratio of own earnings to joint earnings

Mean Ratio of own earnings to joint earnings

In order to assess earnings inequalities within couples, for each person the amount that they contribute to joint earnings has been expressed as a proportion or ratio of joint earnings. A ratio of 0.5 means that the person contributes one half of joint earnings, ie both partners are equal earners. In Figure 12.1, the mean ratio of own earnings to joint earnings has been calculated for each five-year age group. Figure 12.1 shows that men still earn about two thirds of couples’ joint earnings across all age groups, varying from 64 per cent in their late twenties and late fifties to more than 72 per cent in their early forties. The late twenties represents the most equal period in the lives of men and women—before gender pay gaps are much in evidence and before many women have begun having children—although even at this stage women are not equal earners, providing on average only a third of joint earnings. The late fifties is likely to represent a time of early or semi-retirement for men while their (often younger) partners are still working, possibly full time, as children are likely to have grown up or left home; at this stage women make their highest mean contribution, of just over 40 per cent. Between these ages, however, women’s contributions dip, to less than 30 per cent in their late thirties. While the increased equality of couples in their twenties may reflect a cohort change in attitudes to earnings equality, it is difficult to discern any evidence here of cohort changes in the balance of earnings among couples over different age groups. There is little difference between partnered women in their twenties and in their forties. The U-shaped pattern implies strongly that the dip from age 30 to age 44 is related to child-rearing and not cohort changes. Women aged between 30 and 35, for example, are probably not materially different in education, training, attitude, and partnering patterns to women aged between 20 and 29. Women

.42 .40 .38 .36 .34 .32 .30 .28 .26 .24 20-24

25-29

30-34

35-39 40-44 Age Group

45-49'

50-54

55-59

Note: Y axis for men from 0.6 to 0.74; Y axis for women from 0.24 to 0.42 Source: GHS 2000/1 and 2001/2; author’s analysis; Couples with no earnings between them have been excluded. Figure 12.1: Mean ratio of own to joint earnings across age groups. Partnered men and women aged 20 to 59

Pension Accumulation and Gendered Household Structures 263 MOTHERHOOD AND INEQUALITY WITHIN COUPLES

.5

.4

Maternal status never had a child .3

child 0 - 5 child 6 - 15

.2 20-29

.5

ch 16+ home or gone 30-39 40-49 Age group

50-59

Women 2nd quintile: joint earnings .5

.4

Maternal status never had a child

.3

child 0 - 5 child 6 - 15

.2 20-29

ch 16+ home or gone 30-39 40-49 Age group

Women

Women

Third quintile: joint earnings

Fourth quintile: joint earnings

.4

Maternal status

.3

never had a child child 0 - 5

.2

child 6 - 15 .1 20-29

Mean Ratio of own earnings to joint earnings

Women Highest quintile: joint earnings

ch 16+ home or gone 30-39 40-49 Age group

50-59

Mean Ratio of own earnings to joint earnings

Mean Ratio of own earnings to joint earnings

Mean Ratio of own earnings to joint earnings

As the interaction with age just shown suggests, and as Figure 12.2 makes explicit, these persistent-looking patterns of earnings inequality between partners are to a large extent the result of child-care and allied household responsibilities that women still take on to such a substantial degree in the UK. Here, the extent of women’s contributions to joint household earnings is shown by joint earnings

50-59

.6 .5 .4

Maternal status never had a child

.3

child 0 - 5 .2 .1 20-29 30-39 AGE10YR

child 6 - 15 ch 16+ home or gone 40-49

50-59

Mean Ratio of own earnings to joint earnings

Women .7

Lowest quintile: joint earnings

.6 .5

Maternal status

.4

never had a child

.3

child 0 - 5

.2

child 6 - 15

.1 20-29

ch 16+ home or gone 30-39 40-49 Age group

50-59

Note: Scale on y axis varies , Source: GHS 2001 & 2002, author s analysis; Couples with no earnings between them have been excluded. Figure 12.2: The impact of motherhood on earnings’ equality within couples. Women aged 20 to 59 living with a spouse or cohabiting

264 Debora Price quintile and by maternal status in 10-year age groups. Joint quintiles are shown to establish whether different patterns of dual earnings emerge across the joint income distribution—that is, does it make a difference if the woman is in a highearning or low-earning couple? In general, it makes surprisingly little difference. The solid line in each graph denotes women who have never had a child. The next two lines represent the age group of the youngest child in the family unit: either 0–5 or 6–15. The final line shows that the youngest child is either still at home but over 16 or has left home—that is, these are women who have had children, but none of their children, save in exceptional circumstances, could be considered as in need of daytime care. The impact of motherhood on gender inequality among couples across all levels of joint earnings is clearly shown. In each graph the solid line floats well above the others. Women who have never had children (which will, of course, be a diminishing group with age) are contributing between 40 and 50 per cent of joint earnings on average within each age group, and an even higher percentage in the lowest quintile. Whilst clearly not representing complete equality between the sexes, it is, in historical context, a reasonably close approximation. Interestingly, there is no substantive dip among childless couples with age, which would suggest that changes in society in the sense of expectations of dual earning within couples are impacting as much on the financial relationships between mid-life couples as on younger couples. The effect of having children depends to some extent on which earnings bracket a couple falls into, but the broad picture is similar across all five quintiles. Women whose youngest child is under six contribute on average between 24 per cent and 29 per cent to joint earnings in all age groups in the highest two quintiles, while women in mid-earning couples (3rd quintile) in their forties with young children are contributing just 13 per cent to joint earnings. In the lowest two quintiles for women, the contributions of women in their twenties and thirties with a very young child are reasonably constant, at around 20 per cent, but for the lowestearning couples women’s contribution rises sharply if they are in their forties, reaching 41 per cent. Despite being relatively ‘equal’ compared with women in better-off households, this reflects the low earnings of their partners, rather than their own higher earnings. Women in their forties with their youngest child aged 0–5 in the lowest joint earnings quintile were contributing on average just £90 a week, or £4,680 per year, from earnings to household finances in 2002. Thus, women with very young children in the two highest joint earnings categories are contributing a slightly higher proportion on average than women in other categories, which is likely to be a factor in the classification of the couple into those higher earnings brackets in the first place. But they are not anywhere near equal contributors, and the differences in proportions among the various women shown here are surprisingly slight. The low average percentage of joint earnings of women with a child under six years old is, of course, in part a reflection of the higher likelihood that mothers of such young children will be out of the paid labour market (p = 0.345), but neither are those who are in the labour market, on the whole, high

Pension Accumulation and Gendered Household Structures 265 relative contributors (data not shown). Although overall there is some improvement towards equality where the youngest child in a family is at school (aged 6–15), this change is barely discernible in some earnings brackets. The impact of ever having had a child is more difficult to interpret. The lowest quintile apart, women who have ever had a child do not in any age group contribute more than 35 per cent to joint earnings, and there is no evidence here to suggest change for younger cohorts. This could be because of the effects on human capital of years out of the labour market and/or in part-time work, or a gendered accommodation in the household of work, housework, and care in a balance that couples are reluctant to disturb if finances do not require it. The highest quintile stands out for its persistent inequality for mothers of all ages and with children of all ages; the lowest quintile for the degree of earnings equality among couples where a woman is over 40, regardless of the ages of her children. This is most likely a ‘needs must’ situation—with relatively low-earning women partnered with men who are not earning or are themselves low earners. The result of this analysis is that while couples without children are on average reasonably close to each other in earnings, substantial inequalities persist among couples of working age across the economic distribution where they have children, or have ever had children. Women who have ever had a child do not, in any age group or any part of the earnings distribution on average contribute more than about third to joint earnings, unless they and their partners are among the lowest earners. The most equal-earning couples with children are therefore those in the lowest earnings quintiles, rather than the high-powered dual-earning couples prevalent in media discourses. This raises the question of whether higher educational qualifications lead to greater egalitarianism among couples, or whether this too is a myth.

EDUCATION

Policy on equality, driven by the liberal agenda of equality of opportunity, has largely focused on reducing discrimination in education and at work. The theory is that if women are as well qualified as men and workplace discrimination is reduced, then equality between the sexes will follow. But, as just shown, inequality between the sexes within partnerships remains manifest, especially if the parties have children. Lack of available affordable child care and ungenerous maternal/ paternal leave policies could in theory lead to an earnings sacrifice by either parent, but in practice, it seems to be women’s earnings that are sacrificed. The question arises whether this inequality is evident for the most educated women too. Research has shown that new opportunities, particularly for reduced economic dependence, are largely available to highly educated, middle-class, partnered women, and that there is increasing divergence between childless women and mothers (Dex, Joshi and Macran, 1996; Crompton and Harris, 1998; Warren, 2000; Elliott, Dale and Egerton, 2001; Ginn, 2003). Davies, Joshi and Peronaci

266 Debora Price (2000) and Rake et al (2000) have suggested that while the loss of gross earnings from motherhood over a lifetime remains substantial, these losses are mostly borne by less-educated women. Ginn and Arber (2002), however, have challenged these assumptions in so far as they relate to highly qualified mothers, showing that in the mid 1990s fewer than half of graduates who have children were in full-time employment, regardless of age, cohort, and the ages of their children, with sizeable commensurate financial losses. Indeed, in an institutional and cultural setting that encourages it, patterns of educational homogamy in marriage and partnership can theoretically lead to highly educated women being more likely than other women to sacrifice a career for child-rearing, since they are more likely to have high-status, high-earning partners who can afford to maintain the family. Figure 12.3 shows the relationships between women’s education level, their age group, and the degree of earnings inequality in their relationships. The clusters of lines representing different age groups are situated fairly closely together in each case. This means that a woman’s own age is a less important determinant of inequality than the age of her youngest child. The impact of educational attainment is shown by the way the lines in each cluster tend to slope upwards from left to right, with the left-most data points representing those with no qualifications, and the right-most data points women with the highest qualifications, such as degrees. The extent of equality of earnings in relationships varies little according to educational attainment for women who have never had a child, with such women generally contributing about 40 to 50 per cent of family earnings, whatever their age and educational attainment, indicating a relatively high degree of earnings homogamy among partners without children. Looking at those with a young child under the age of six in the family, in each age range higher education means women are more equal than those with lower education. However, here even women with degree-level education are contributing only around 30 per cent of joint earnings—while their children are young they earn on average less than half of the earnings of their partners. With children this young, educational attainment has a notably stepped impact on equality within the household—higher education implies greater equality but not even near-equal earnings. Women with degree-level education are also the most equal in terms of earnings within their partnerships when they have children at school, but the mean ratio of their own to joint earnings is still only about a third. They are also the most equal controlling for age group when their children have grown up, but it is only among degree-educated mothers in their fifties that virtual equality of earnings is reached. Moreover, it is only degree-level education that has this effect once children are of school age. While women with no qualifications at all are always, on average, lower relative contributors to household earnings, among other women with school-age or older children, qualifications make little difference to equality, whatever the age group. In summary, women’s degree-level education brings greater earnings equality to relationships, but not as much as one might think. Degree-educated women with

0

0.1

0.2

0.3

0.4

0.5

Level 3 (e.g. A'levels)

Level 2 (e.g. O-levels)

Level 1 (basic)

Never had a child

Level 3 (e.g. A'levels)

Level 2 (e.g. O-levels)

Level 1 (basic) Youngest child 0 - 5

50 - 59

40 - 49

30 - 39

20 - 29

Level 3 (e.g. A'levels) Level 2 (e.g. O-levels) Level 1 (basic)

Youngest child 6 - 15

Level 4/5 (highest degree etc) Level 2 (e.g. O-levels) Level 1 (basic) No qualifications

Youngest child 16+ (home or gone)

Level 3 (e.g. A'levels)

Level 4/5 (highest degree etc)

No qualifications

Level 4/5 (highest degree etc)

No qualifications

Level 4/5 (highest degree etc)

No qualifications

Figure 12.3: Mean ratio of own earnings to joint earnings by maternal status, educational qualifications and age group. Partnered women aged 20 to 59

, Source: GHS 2001 & 2002, author s analysis; those in partnerships where neither party earns are excluded; also excluded are cells where n < 15, and those with unclassifiable or unknown qualifications.

Mean ratio of own to joint earnings

0.6

Pension Accumulation and Gendered Household Structures 267

268 Debora Price young children and children at school experience marked inequality in earnings at home, whatever their age, and it is only if they do not have children, or their children have grown up, that this changes. Educational attainment and age are also not the most important drivers of earnings inequality for women with educational attainment short of a degree. The most important influencing factor is once again the presence of dependent children in the household, or having ever had a child. PARTNERSHIP, DEPENDENCY AND PENSION PROVISION

Mothers, therefore, remain particularly dependent on their partners for household income. The question then arises whether this economic dependency matters for participation in pension schemes. Figure 12.4 shows the percentage of partnered women aged 20 to 59 who are participating in third-tier pensions according to the degree of earnings inequality in their relationship. This shows that women who earn less than a fifth of joint earnings (a third of partnered women) are the least likely to contribute to pensions, with barely 10 per cent making such contributions. Equal and better earners are more likely to contribute, but only just over half of the few women who are the primary earners in their relationships (earning more than 80 per cent) do so. 90

Percent in third tier pensions

80 70 60 50 40 30 20 10 0 0% < 20% (35%)

20% - 40% (28%)

40%-60% (24%)

60%-80% (4%)

80% - 100% (9%)

All (100%)

Degree of earnings equality

Figure 12.4: Percentage of partnered women aged 20 to 59 participating in third-tier pension schemes according to degree of earnings equality. Brackets show distribution of partnered women

Pension Accumulation and Gendered Household Structures 269 It is not necessarily the case, however, that these patterns are related to the degree of economic dependency in the relationship. They might be a result of the amount that the individual women are earning, or how much income comes into the household, since we have shown that it is in the lower quintiles of earnings that women are most equal. It may also be related to the lower educational qualifications of these relatively low-earning women, or to the fact that low-earning women might on average have younger children. To elucidate the mechanisms by which the inequality within partnerships is related to making provision for additional private pension during the working life, multivariate analysis is needed, controlling for a number of possibly influencing factors simultaneously. Table 12.1 presents the results of a multivariate logistic regression analysis for partnered women aged 20 to 59. The analysis in this table shows how earnings inequality is related to third-tier pension coverage, while taking into account all of these other variable factors in each model. The results show the odds ratios for being in this privileged category of those with pension coverage relative to a reference category for which the odds have been defined as 1. The number in the table tells you, relative to the reference category, how much greater the odds are that a person with that characteristic has a third-tier pension compared with the reference category, while the other attributes in the model are controlled. In Model 1, the odds ratios are shown according to the extent and direction of earnings inequality within a relationship, with reference to the category ‘40%–60%’, who for these purposes are considered ‘equal’ earners, while age group is controlled. So we see that the odds of a woman who earns 20 to 40 per cent of joint earnings contributing to a third-tier pension are only approximately half (0.49) the odds of an equal earner, controlling for age group; and for someone contributing 0 to 20 per cent of joint earnings, the odds of contributing to a third-tier pension are only four per cent of the odds for an equal earner—the odds for an equal earner are 25 times higher. Moving from left to right across this table shows how the impact of inequality of earnings on third-tier pension contributions is affected by other variables. Model 2 also controls for individual earnings, and Model 3 for joint earnings. In Model 4, maternal status is controlled, together with educational qualifications. This is to test whether, apart from their influence on individual earnings, joint earnings, and degree of inequality within partnerships, these variables have any independent impact on participation in pension provision, and whether they affect the association between inequality and pension scheme participation. The primary question of interest here is the effect of gender inequality of earnings within couples on participation in third-tier pension schemes. Although these models contain much other information, this analysis will comment only on the coefficients across models for the first category, that is the percentage of joint earnings (or the degree of earnings inequality). As indicated in these tables, almost two thirds of women contribute less than 40 per cent to joint earnings, and fewer than one in ten more than 80 per cent. The odds ratios in this category after all these controls (that is, for Model 4) are presented graphically in Figure 12.5.

270 Debora Price Table 12.1: Odds ratios for contributing to additional pensions. Partnered women aged 20 to 59 Model 1

Model 2

Model 3

Model 4

Percentage of joint earnings 0 < 20% [35% of women] 20% < 40% [28% of women] 40% < 60% [24% of women] 60% < 80% [4% of women] 80% < 100% [9% of women]

*** 0.04*** 0.49*** 1.00 1.15 0.40***

** 0.76* 1.13 1.00 0.85 0.76*

*** 0.35*** 0.81* 1.00 1.17 1.63**

*** 0.34*** 0.80* 1.00 1.14 1.51**

Age group 20–29 30–39 40–49 50–59

*** 0.43*** 1.00 1.13 0.82*

*** 0.41*** 1.00 1.05 1.06

*** 0.42*** 1.00 1.04 1.14

*** 0.40*** 1.00 1.10 1.20

*** 1.00 0.53*** 0.31*** 0.14*** 0.02*** 0.00~

*** 1.00 0.67* 0.57** 0.36*** 0.10*** 0.00~

*** 1.00 0.70 0.64** 0.43*** 0.12*** 0.00~

*** 1.13 1.00 0.60*** 0.46*** 0.18***

*** 1.06 1.00 0.63*** 0.49*** 0.20***

Earnings quintile: own earnings Highest 2nd 3rd 4th Lowest No earnings Earnings quintile: joint earnings Highest 2nd 3rd 4th Lowest Maternal Status Never had a child Youngest 0–5 Youngest 6–15 Youngest over 16 (home or gone)

* 1.00 1.22 0.91 1.20

Educational qualifications Level 4/5- highest (degree etc) Level 3 (e.g. A-levels) Level 2 (e.g. O-levels) Level 1 (basic) No qualifications Other or unknown

*** 0.94 0.80 1.00 0.61*** 1.43*** 0.69*

-2LL Change in chi square DF

5540 1741 3

4407 1133 5

4333 74 4

4275 58 8

***p < 0.001; **p < 0.01; *p < 0.05 (no asterisk = not statistically significant at 5%) ~odds of contributing to a pension defined as 0 Source: GHS 2000/1 and 2001/2, author’s analysis, Excludes couples where neither earns.

Pension Accumulation and Gendered Household Structures 271 1.6

Odds ratios (ref: equal earners)

1.4 1.2 1 0.8 0.6 0.4 0.2 0

0% < 20% (35%)

20% - 40% (28%)

40%-60% (24%)

60%-80% (4%)

80% - 100% (9%)

Degree of earnings equality

Figure 12.5: Odds ratios for participation in third-tier pension schemes according to degree of earnings inequality in relationship, partnered women aged 20 to 59, after controlling for age group, own earnings, partner’s earnings, maternal status and educational qualifications

Female breadwinners are more likely to be low earners. This is shown by the dramatic increase in the odds ratios for women providing over 80 per cent of joint earnings from 0.4 in Model 1 to 0.76 in Model 2, when the strong association between low earnings and low third-tier pension scheme participation is taken into account. Women providing 80 per cent plus of earnings as well as being low earners themselves are also much more likely to be in the low end of the joint earnings distribution. Thus it is only when both personal earnings and joint earnings are controlled that we see the odds ratio for the pension contributions of ‘breadwinning’ women jump from 0.76 to 1.63 (p < 0.01), relative to an equal earner. Model 3 shows that the degree of earnings inequality in a relationship has a significant impact on whether women contribute to pensions. Women who are more financially dependent are less likely to contribute to additional private pensions. For women earning 0–20% of joint earnings, the odds are only a third of those for equal earners, and for those earning between 20% and 40% of joint earnings, the odds ratio is 0.81. All other things being equal, women who have assumed the role of major breadwinner are more likely than equal earners to make additional pension provision, conforming to a breadwinning norm. But all other things are not equal—the starting odds ratio for women in couples where they are the major breadwinners is only 0.4 because these effects are counterbalanced by their higher probability of low earnings and low joint earnings where third-tier pension

272 Debora Price provision is very unlikely. The additional statistical controls of maternal history and educational qualifications do not affect these relationships. These results suggest that financial dependency within a partnership is normatively associated with lack of third-tier pension accumulation. Being a ‘second earner’ has implications beyond immediate access to financial resources—it is also related to relying on a partner or the state for pension provision. This is important given the high risk of partnership dissolution or widowhood over the life-course.

PENSIONS AND PARTNERSHIP STATUS

Within partnerships, where there are children, women are likely to be financially dependent and to have low rates of contribution to third-tier pensions. Yet they risk separation and divorce, and living alone in later life. It is important to establish how likely women are to contribute to third-tier pensions once partnerships have ended, and this is the final question addressed in this chapter. Unravelling whether it is partnership status or children that influence pension scheme participation is complex. Women in different partnership statuses are not equally likely to have children at home, and children are an important factor in women’s earnings and pension accumulation. Also, women in different partnership statuses are likely to be of different ages, and age is also strongly associated with pension accumulation. This is because pensions become more salient to people as they age. Table 12.2 shows the percentage with children in each marital status, and the mean age of that subgroup, among women aged 20 to 59. The age profiles of divorced lone women and divorced women who are cohabiting are very similar, but divorced lone women are more likely to have a child under 16 (48 per cent as opposed to 42 per cent) and also more likely to have a child who is over 16 (13 per cent as opposed to 8 per cent). Divorcees with children have lower prospects of repartnering (Haskey, 1999; Lampard and Peggs, 1999). There are substantial age difference between those in second marriages, cohabiting divorced women, and cohabiting women who have never been married—in particular, never-married cohabitants are much younger on average. Never-married cohabitants are more likely to be childless (p = 0.63), but, notably, more than a quarter have a child under six at home. A quarter of never-married lone women have a child under 16 to look after—a smaller proportion than in any other subgroup, but substantial nevertheless. More never-married lone mothers have a child under six to care for than do divorced lone mothers. The percentage of men and women in third-tier pension schemes in each partnership status is shown in Table 12.3. Although the variation among women is not that great, lone women are much less likely to have current pensions than partnered women (27–35 per cent, compared with 38–45 per cent), and among partnered women, divorced cohabitants are the most likely to be contributing (45 per cent).

30

45

34

27

30

Mean age

849

100%

1

10

27

63

%

29

42

34

29

29

Mean age

Never married, cohabiting

5,131

100%

17

25

24

35

%

42

52

40

32

39

Mean age

In first marriage

341

100%

13

28

28

31

%

39

50

38

31

36

Mean age

Sep from first marriage, lone

650

100%

13

35

13

40

%

44

51

38

33

44

Mean age

Div from first marriage, lone

281

100%

8

22

20

49

%

42

48

39

33

41

Mean age

Div from first marriage, cohabiting

Source: GHS 2001 & 2002, author’s analysis. Notes: Mothers whose dependent children live elsewhere and those looking after others’ children have been excluded.

1,746

3

All children 16+ (home or gone)

n=

12

Youngest 6–15

100%

14

Youngest 0–5

All

71

%

Never married, lone

Never had a child

Women

Table 12.2: Percentage of population with dependent children and mean age by selected marital statuses

813

100%

16

25

16

43

%

46

51

42

37

56

Mean age

In second marriage

Pension Accumulation and Gendered Household Structures 273

274 Debora Price Table 12.3: Percentage in third-tier pension schemes according to partnership status, men and women aged 20 to 59 Percentage in third-tier pensions

Never married, lone Never married, couple First marriage Separated (1st mar), lone Divorced (1st mar), lone Divorced (1st mar), couple Widowed (1st mar), lone Second marriage Complex history (2+ mar ended)

Women

Men

35 40 40 29 33 45 27 38 35

36 52 65 58 40 58 47 56 39

Source: GHS 2001 & 2002, author’s analysis. Note: Separated men and women who are cohabiting, and cohabiting widows and widowers are excluded for small numbers; those with complex histories include both lone men and women, and couples, who have lived through the breakdown of at least two marriages.

To establish the extent to which this is related to the legal marital status and marital history of any women, as opposed to age or the presence of children, once again, multivariate analysis is needed. Table 12.4 presents odds ratios resulting from a logistic regression analysis, again predicting whether the respondent participates in a third-tier pension. The odds ratios presented compare the odds of participation for women in other marital statuses to those in a first marriage, where the odds have been set to 1. Age group, fertility history, gross earnings quartile, social class, and highest educational level have been controlled with dummy variables, and again odds ratios are presented compared with the reference group (age 40–49, having a child age 0–5 in the family, lowest quartile, social class IIIM, and Level 3 (one or more A-levels or equivalent) in education). Nested models are presented, with the predictor variables added one by one to see the effect at each level on the association between partnership status and pension scheme participation. The first model shown, Model 1, controls for age group and fertility history. This shows the importance of fertility history, with the odds of contributing to a third-tier pension 3.3 times higher for a woman who has never had a child than a woman who has a pre-school child at home, and twice as high as for a woman whose youngest child is now 16 or over (3.3 compared with 1.6), once age and partnership status are controlled. Similarly, the low participation of women at the extremes of the age range is highlighted, with the odds of contributing to a thirdtier scheme for women in their twenties and fifties about half the odds for those in the middle age-ranges (30–49), controlling for children.

Pension Accumulation and Gendered Household Structures 275 Table 12.4: Odds ratios for participation in third-tier pension schemes. Women aged 20 to 59 n=

Model 1

Model 2

Model 3

Model 4

Partnership status Never married, lone Never married, couple First marriage Separated (1st mar), lone Divorced (1st mar), lone Divorced (1st mar), couple Widowed (1st mar), lone Second marriage

1,514 771 4,534 284 581 260 139 735

*** 0.67*** 0.92 1.00 0.65** 0.73*** 1.02 0.56** 0.86

*** 0.63*** 0.65*** 1.00 0.57** 0.64*** 0.79 0.80 0.80

*** 0.64*** 0.66*** 1.00 0.57** 0.66*** 0.81 0.85 0.81*

*** 0.64*** 0.66*** 1.00 0.58*** 0.67*** 0.83 0.87 0.82

Age group 20–29 30–39 40–49 50–59

1,835 2,725 2,372 2,237

*** 0.49*** 1.00 1.13 0.62***

*** 0.47*** 1.00 1.17* 1.15

*** 0.46*** 1.00 1.17 1.12

*** 0.45*** 1.00 1.17* 1.18

Fertility history Never had a child Dependent child 0–5 Dependent child 6–15 Child 16+, home or gone

2,818 1,574 2,113 2,664

*** 3.33*** 1.00 1.13 1.65***

*** 0.86 1.00 0.66* 0.72

** 0.91 1.00 0.71** 0.80

*** 0.93 1.00 0.72* 0.86

Gross earnings, £ per week Highest quartile Second quartile Third quartile Lowest quartile No earnings

934 1,395 2,065 2,268 2,507

*** 26.36*** 12.50*** 5.34*** 1.00 †

*** 17.98*** 9.45*** 4.71*** 1.00 †

*** 16.63*** 8.99*** 4.67*** 1.00 †

Social Class I II IIINM IIIM IV V

281 2,664 3,282 786 1,654 502

*** 2.65*** 2.40*** 2.01*** 1.00 1.18 0.79

*** 2.15*** 2.06*** 1.92*** 1.00 1.21 0.89

Highest Educational Level Level 4/5 (highest) Level 3 Level 2 Level 1 No qualifications Unknown

2,499 1,355 2,233 687 1,819 576

*** 1.26* 1.00 1.04 0.86 0.69*** 0.84 (continued)

276 Debora Price Table 12.4 (Continued) n= -2LL change in -2LL df Significance

Model 1

Model 2

Model 3

Model 4

11,793 535 14 ***

7,262 4,531 4 ***

7,142 120 5 ***

7,109 33 5 ***

Source: GHS 2001 & 2002, author’s analysis. Notes: *p < 0.05; **p < 0.01; ***p < 0.001 (no asterisk = not statistically significant at 5%); † – probability of third-tier pension provision defined as zero; Same sex cohabitees, separated women cohabiting, and widows cohabiting have been excluded due to small numbers; those with complex marital histories whereby more than two marriages have ended are not shown due to conceptual difficulties in analysis.

Model 1 shows an apparently clear distinction between women who live alone, and those who live with a partner. Like for like in maternal status and age group, lone women (never-married, separated, divorced and widowed) are much less likely than married women to have third-tier pensions (odds ratios of 0.67, 0.65, 0.73 and 0.56 respectively), whereas there are no statistically significant differences3 between women who are living with a partner, whether never married, in a first marriage, divorced, or in a second marriage. The apparently higher rate of contribution to third-tier pensions of divorced cohabiting women in Figure 12.5 is shown to be a function of their lower likelihood of having children, and their higher likelihood of being in their forties (peak age for pension accumulation for women); once these variables are controlled they are shown in fact to be less likely than married women to be making contributions to third-tier pensions. When earnings are controlled (Model 2), the very great difference that earnings make to third-tier pension scheme participation is obvious, with the odds for a woman in the highest quartile of earnings 26 times higher than for a woman in the lowest quartile. Controlling for earnings explains the differences in odds ratios for women in their fifties, for women who have never had a child and those with grown-up children. It does not explain the lower odds for women in their twenties, confirming that this is an age-related factor. After controlling for earnings, it is shown that women with a school-age child are the least likely to make current contributions to third-tier pension provision. Controlling for earnings (Model 2) serves to accentuate the distinctions between women of different partnership statuses. Despite their low pension contributions, never-married women, divorced women and women in second marriages are all on average higher earners than women in first marriages, after controlling for age and children. Thus when earnings are controlled, the odds ratios for all these groups of women contributing to pensions compared with married women fall 3

At 5%.

Pension Accumulation and Gendered Household Structures 277 even lower than in Model 1. For lone separated women, the odds ratio is 0.57, for lone divorced women 0.64 and for lone never-married women, 0.63. Controlling for earnings also shows that, like for like on other controlled factors, cohabiting women who have never been married are also much less likely than married women to have third-tier pensions, with an odds ratio of only 0.65. The differences between odds for married women, cohabiting divorced women, and women in a second marriage remain statistically non-significant. It might be thought that the lower contribution rates of lone mothers could be explained by the known correlation between lone motherhood and lower social class and educational attainment (Rowlingson and McKay, 2005). However, class differences and educational differences between women in different marital statuses do not account for these findings. As Models 3 and 4 show, controlling for social class and for highest educational qualifications does not alter the coefficients for the partnership status variables, which are in fact fairly robust from Model 2 through to Model 4. This analysis shows that the impact that children have on women’s earnings— both while the children are dependent and afterwards, and whatever the mother’s marital status—is extremely important in pension accumulation. It also suggests that it is important to think of the ways in which having children might affect women’s pension accumulation apart from a simple effect on earnings—if mothers return to paid work when their children are at school, social norms about how mothers spend money on child care and children while they are dependent (or beyond) may be diverting women’s earned income towards children at the expense of their pensions, thus explaining the robust lower odds of pension contribution for women with children of school age. Women’s exit from paid labour and lower earnings in their fifties—likely to be for social reasons as well as labour market reasons—has a further deleterious effect on their pensions. Partnership status and partnership history are also important. Cohabiting women who have never married are higher earners than other women, but once earnings are controlled, married and divorced cohabiting women are more likely to have third-tier pension provision. This may mean that there is something about having been in or through a marriage that focuses women’s minds on pension accumulation, or may indicate a selection effect—those women who are more likely to make pension provision are also more likely to marry. However, significantly, compared with women in couples, lone women are the least likely to be contributing to third-tier pension provision despite the importance of doing so for their financial well-being in later life. Once age, motherhood, earnings, social class, and educational level are controlled, lone women are much less likely than those in couples to make pension provision. The reasons are unrelated to their earnings, social class, or educational attainment. Despite the already low rates of pension accumulation among married women, among lone women rates are even lower. Very little is known about what happens to the financial well-being of divorced men and women in the UK after divorce (see Fisher and Low, chapter eleven,

278 Debora Price this volume), but what is known suggests that women emerge from the process with considerable financial burdens. Few fathers pay child support. Estimates vary for the percentage of lone mothers in receipt of any maintenance from the fathers of their children from 23 per cent (Rowlingson and McKay, 2005) to 31 per cent (Marsh and Perry, 2003). Rowlingson and McKay (2005) report that among manual workers the figure is only 16 per cent. These data support the notion that the financial constraints of single living are such as to prevent lone women securing a financial future for themselves through the acquisition of third-tier pensions. The association holds even for those without children, however, suggesting that being part of a couple seems to act as an enabling or normative mechanism for third-tier pensions for women, or that personality factors or economic incentives are associated both with contributing to pensions and being in a couple. This is especially worrying given that on average married women start from a low base. The wide disparity in pension rights as between husband and wives (Field, 2000) led ultimately to government adding to the courts’ discretionary powers on divorce a further power to allocate or ‘share’ pensions between spouses.4 The government declared that pension-sharing was ‘an important step towards security in retirement for women’ (DSS, 1998: Foreword by the Secretary of State). Despite several years of operation, it is little used (Price, 2003). In 2006 there were 13,630 pension-sharing orders made (Ministry of Justice, 2007: 91, table 5.6), approximately one per cent of divorces. If women cannot expect the divorce courts to provide them with a share of accumulated ‘family’ pension provision, then they are likely to emerge from their marriages with no or little pension. These data suggest that if they have not repartnered, they are far less likely than married women to be accumulating thirdtier pensions, particularly if they have children. CONCLUSIONS

The pension structure in the UK demands that for adequate later-life individual income, individuals must accumulate private pensions. Despite forthcoming reforms to the state system, and better financial incentives for many to participate in private pensions, histories of involvement with paid work, earnings, and private pension accumulation will remain crucial in determining pension outcomes for women (Ginn, 2003; Price, 2007, 2008). To the extent that work biographies remain highly gendered, private pension systems will continue to disadvantage women.

4

In England and Wales, by Parts III and IV of the Welfare Reform and Pensions Act 1999. The Act also applies in Scotland, but the legal system for the division of matrimonial property in Scotland is quite different.

Pension Accumulation and Gendered Household Structures 279 However, this chapter has shown that it is not only work biographies but also gendered cultural norms which influence women’s participation in third-tier pension schemes. The UK has a prevalent culture of women becoming ‘second earners’ in relationships once children are born. Large disparities in earning as between mothers and their partners persist in all income and educational strata. Mothers are likely to be dependent on their spouses or partners to provide the bulk of household income. This financial dependency appears to continue once children have grown up, with little evidence so far of cohort changes. The culture has implications for pension provision, since women on low incomes, and in low income households are much less likely to be making thirdtier provision. Financial dependency itself also affects pension provision, even after controlling for income and various other demographic variables. This analysis suggests that women who rely on their partners for income during their working life continue to anticipate that their partners will also provide for them in old age. When partnerships break down, lone separated and divorced women are disadvantaged in pension accumulation. Lone mothers especially do not generally participate in a third-tier pension. Whether single, separated or divorced, lone mothers have lower rates of third-tier pension accumulation than married women, even taking into account likely selection effects such as education and social class. Separated mothers, bearing the financial strains of the immediate aftermath of family breakdown, have the lowest rates of all. After family breakdown, mothers are likely to spend many years as a lone mother, with re-formed partnerships themselves fragile. Multivariate analysis confirmed that divorced women who are cohabiting do not accumulate pensions at any greater rate than other partnered women, thus conforming to the same social and financial patterns of behaviour as married women, and as are prevalent around them in society. Neither lone divorced women nor cohabiting divorced women show the higher rates of pension accumulation after divorce that we might expect to see if they were ‘catching up’ in pension building. If anything, the data suggest strongly that lone divorced mothers are falling further behind. Advocates of gender equality and a reduction in income poverty experienced by older women have argued for a reformed system where pensions are accrued universally regardless of participation in paid labour (often called a ‘citizen’s pension’), paid at a rate that provides at least a poverty-line income, and that will be increased once in payment in line with national average earnings (PPI, 2006). This would reduce the associations between paid work history, earnings history, and pension income and therefore benefit women who have tended to fare poorly in the UK’s employment-based system. Such a reform has now been rejected by government, which, with cross-party support, has committed to a pension system where work-based contributions and private pensions remain important elements for income adequacy in later life. Without substantial changes in gender relations within households and in the policy and labour market structures that support them, mothers who may live alone in later life continue to face a high risk of financial insecurity.

280 Debora Price BIBLIOGRAPHY Arber, S and Ginn, J (1995) ‘The mirage of gender equality: occupational success in the labour market and within marriage’ 46 British Journal of Sociology 21. —— (2004) ‘Ageing and Gender: Diversity and Change’ in Summerfield, C and Babb, P Social Trends 34 (London, The Stationery Office). Crompton, R and Harris, F (1998) ‘Gender relations and employment: the impact of occupation’ 12 Work, Employment and Society 297. Davies, H, Joshi, H and Peronaci, R (2000) ‘Forgone income and motherhood: what do recent British data tell us?’ 54 Population Studies 293. Dex, S, Joshi, H and Macran, S (1996) ‘A widening gulf among Britain’s mothers’ 12 Oxford Review of Economic Policy 65. DSS (Department of Social Security) (1998) Pension Sharing on Divorce: Reforming Pensions for a Fairer Future. Part I: Consultation (London, DSS). DWP (Department for Work and Pensions) (2005) Women and Pensions: The Evidence (London, DWP). —— (2007) Pension Credit Estimates of Take-up in 2005/6 (London, DWP). Elliott, J, Dale, A and Egerton, M (2001) ‘The influence of qualifications on women’s work histories, employment status and earnings at age 33’ 17 European Sociological Review 145. Evason, E and Spence, L (2002) Women and Pensions (Belfast, Equality Commission for Northern Ireland). Field, J (2000) Pensions and Divorce: the 1998 Survey Department of Social Security Research Report No. 117 (Leeds, Corporate Document Services). Ginn, J (2003) Gender, Pensions and the Lifecourse: How Pensions need to Adapt to Changing Family Forms (Bristol, The Policy Press). Ginn, J (2007) ‘Poverty and Financial Inequality in Later Life’ in T Ridge and S Wright (eds), Understanding Inequality, Poverty and Wealth: Policies and Prospects (Bristol, The Policy Press). Ginn, J and Arber, S (2002) ‘Degrees of freedom: do graduate women escape the motherhood gap in pensions?’ 7 Sociological Research Online. Glennerster, H (2006) ‘Why so different? Why so bad a future?’ in H Pemberton, P Thane and N Whiteside (eds), Britain’s Pensions Crisis: History and Policy (London, The British Academy). Haskey, J (1999) ‘Divorce and remarriage in England and Wales’ 95 Population Trends 18. Johnson, P and Stears, G (1996) ‘Should the Basic State Pension be a contributory benefit?’ 17 Fiscal Studies 105. Lampard, R and Peggs, K (1999) ‘Repartnering: the relevance of parenthood and gender to cohabitation and remarriage among the formerly married’ 50 The British Journal of Sociology 443. Lewis, J (1992) ‘Gender and the development of welfare regimes’ 2 Journal of European Social Policy 159. —— (1997) ‘Gender and welfare regimes: further thoughts’ 4 Social Politics 160. Manning, A and Petrongolo, B (2004) The Part-time Pay Penalty (London, Women and Equality Unit, Department of Trade and Industry). Marsh, A and Perry, J (2003) Family Change 1999 to 2001, Department for Work and Pensions Research Report No 180 (Leeds, Corporate Document Services). McKie, L, Bowlby, S and Gregory, S (2001) ‘Gender, caring and employment in Britain’ 30 Journal of Social Policy 233.

Pension Accumulation and Gendered Household Structures 281 Millar, J (2003) ‘Gender, poverty and social exclusion’ 2 Social Policy & Society 181. Ministry of Justice (2007) Judicial and Court Statistics 2006, Cm 7273 (London, Ministry of Justice). O’Connor, J, Orloff, A and Shaver, S (1999) States, Markets, Families: Gender, Liberalism and Social Policy in Australia, Canada, Great Britain and the United States (Cambridge, Cambridge University Press). Orloff, A (2002) Women’s Employment and Welfare Regimes: Globalization, Export Orientation and Social Policy in Europe and North America (Geneva, United Nations Research Institute for Social Development). Peggs, K (2000) ‘Which pension? Women, risk and pension choice’ 48 The Sociological Review 344. Pensions Commission (2004) Pensions: Challenges and Choices. The first report of the Pensions Commission (London, The Stationery Office). Pfau-Effinger, B (1999) ‘The modernisation of family and motherhood in Western Europe’ in R Crompton (ed), Restructuring Gender Relations and Employment: The Decline of the Male Breadwinner (Oxford, Oxford University Press). PPI (Pensions Policy Institute) (2006) Shaping a Stable Pensions Solution: How Pension Experts would Reform UK Pensions (London, PPI). Price, D (2003) ‘Pension sharing on divorce: the future for women’ in C Bochel, N Ellison and M Powell (eds), Social Policy Review 15: UK and International Perspectives (Bristol, The Policy Press). —— (2006) ‘Why are older women in the UK poor?’ 7 Quality in Ageing 23. —— (2007) ‘Closing the gender gap in retirement income: what difference will recent UK pension reforms make?’ 36 Journal of Social Policy 1. —— (2008) ‘Towards a new pension settlement? Recent pension reform in the UK’ in T Maltby, P Kennett and K Rummery (eds), Social Policy Review 20: Analysis and Debate in Social Policy (Bristol, The Policy Press). Rake, K, Davies, H, Joshi, H and Alami, R (2000) Women’s Income over the Lifetime (London, Women’s Unit, Cabinet Office). Rowlingson, K and McKay, S (2005) ‘Lone motherhood and socio-economic disadvantage: insights from quantitative and qualitative evidence’ 53 The Sociological Review 30. Smithson, J, Lewis, S, Cooper, G and Dyer, J (2004) ‘Flexible working and the gender pay gap in the accountancy profession’ 18 Work, Employment and Society 115. Walby, S and Olsen, W (2002) The Impact of Women’s Position in the Labour Market on Pay and Implications for UK Productivity (London, Women and Equality Unit/Department for Trade and Industry). Ward, C, Dale, A and Joshi, H (1996a) ‘Combining employment with childcare: an escape from dependence?’ 25 Journal of Social Policy 223. —— (1996b) ‘Income dependency within couples’ in L Morris and E Lyon (eds), Gender Relations in Public and Private: New Research Perspectives (Basingstoke, Macmillan). Warren, T (2000) ‘Women in low status part-time jobs: a class and gender analysis’ 4 Sociological Research Online 113. —— (2001) ‘Divergent female part-time employment in Britain and Denmark and the implications for gender equity’ 49 The Sociological Review 548. —— (2003) ‘Class- and gender-based working time? Time poverty and the division of domestic labour’ 37 Sociology 733. —— (2004) ‘Working part-time: achieving a successful “work–life”’ balance?’ 55 The British Journal of Sociology 99.

282 Debora Price Women and Equality Unit (2005) Women and Men in the Workplace (London, Women and Equality Unit/Department for Trade and Industry). Woods, L, Makepeace, G, Joshi, H and Dolton, P (2003) ‘The world of paid work’ in E Ferri, J Bynner and M Wadsworth (eds), Changing Britain, Changing Lives. Three Generations at the Turn of the Century (London, Institute of Education).

13 Rational Decision-making and Intimate Cohabitation ANTONY W DNES

INTRODUCTION

I

n this chapter, I discuss the approach taken in ‘law and economics’ studies of the family to questions concerned with intimate unmarried cohabitation. I refer henceforth simply to ‘cohabitation’, in line with the practice in standard legal writing, accepting that most readers will realise that the discussion relates to lovers, not grandmothers. Work in law and economics is simplistically associated, in many minds, with models and applied work in which humans are expected to behave purposively and may display rational intent; in fact, much of the work actually recognises strategic interaction, which is not conventional rational decisionmaking. I propose to analyse several strands of work focused on marriage, making some comparisons with cohabitation along the way, and thereby illustrating particularly useful contributions made by the law and economics approach. Broadly, these strands include the search for a satisfactory theory of marriage, compared with cohabitation, the significance of empirical studies of trends in marriage and cohabitation, and policy design—approached as a special case of the economic analysis of regulation. My hope is to show that such rationality-based studies yield fruitful results that could not be generated by alternative approaches to family law. I begin, however, by distinguishing between normative and positivistic (often referred to as ‘positive’) studies in economics, keeping in mind the family area of application.1 NORMATIVE AND POSITIVISTIC ANALYSIS

Many questions concerning intimate cohabitation can be beneficially analysed using approaches based on the enlightened thought2 characteristic of modern 1 My practical focus on the adult-relationship side of family law reflects the balance of my own interests. Economists have also contributed to debates on surrogacy, adoption, domestic violence, child abuse, inheritance and the treatment of elders (Dnes, 2005: ch 9). 2 I mean enlightened in the sense of its being a product of the Enlightenment and the development of scientific method emphasising measurement and testing of hypotheses.

286 Antony W Dnes studies in law and economics. It is essential in carrying out such exercises to distinguish between normative analysis that considers intelligently designed social policy and positivistic3 analysis that predicts how people might respond when faced with changes in socio-economic or legal variables. We should bear in mind that economic benefits come in many shapes and sizes, are defined as elements that contribute to a person’s sense of well-being, and are not limited to pecuniary effects.4 To some extent, what follows gives an insight into the economist’s view of rationality, which really requires little more than people preferring more over less of some beneficial quantum. The qualification is important because, beyond this extent, there are many studies, particularly of human happiness (Layard, 2005), that recognise the influence of apparently irrational behaviour on well-being. From my own observations, mixing with people across different academic camps, there is often a striking need to distinguish between (i) the explanation of what is happening to the family and (ii) an acceptable basis for analysing the desirability of particular policy approaches toward the family: distinguishing positivistic studies from normative analysis helps enormously in this regard.

Normative Analysis Generally, normative analysis formulates objectives and recognises constraints. Economists draw on the principles of welfare economics, in the sense of figuring out how people are made better or worse off by policy changes, which, in relation to jurisprudence, reflects an underlying utilitarianism (Dnes, 2005). Welfare economics could use alternative perspectives from jurisprudence, such as Rawls’ (2005) egalitarian focus, or Nozick’s (1974) individualist focus, although often this would make no difference to the result. Utilitarianism, in its modern guise of welfare economics, and even possibly simplified to wealth maximisation (Posner, 1984: ch 1), provides an analytical system for the analysis of the settling-up aspects of relationship breakdown: it broadly tells us the rule on which the parties would have agreed had they thought about dissolution ahead of time. From this perspective, the parties should choose a settlement rule maximising their joint welfare and leaving each at least as well off as they would be if they had never married. In relation to settling up, utilitarianism does not generate the problems with which it is associated in other contexts.5 The use of ‘what the parties might have chosen’ as a guide to settlements has surfaced in the House of Lords’ continuing discussion of ‘imputation’ and 3 Richard Lipsey popularised the term ‘positive economics’ back in the 1960s: see Lipsey (2008). I prefer the term ‘positivistic’, which is historically more accurate. Modern economics comes straight to us from the Scottish Enlightenment, via Adam Smith, David Ricardo, Jeremy Bentham, and John Stuart Mill. 4 A focus on pecuniary effects would give a kind of ‘law and accountancy.’ 5 English jurisprudence emphasises the context-dependency of the reasonable application of systems of moral philosophy, of which utilitarianism is one example.

Rational Decision-making and Intimate Cohabitation 287 ‘inference’ of parties’ intentions regarding the ownership of the shared home by spouses and cohabitants (and in other ‘domestic’ contexts), most recently in Stack v Dowden.6 The distinction between imputation and inference is blurred by the majority in Stack, who acknowledge that there is a distinction but then go on to use the two terms interchangeably. Lord Neuberger’s dissent in Stack does distinguish between inference and imputation7 Inference entails discerning the parties’ actual (albeit unexpressed) intentions, which is a positivistic exercise in keeping with the idea of ascertaining the true nature of the parties’ property rights. In contrast, imputation entails ascribing to the parties the intentions that we think they would have had, had they thought about the issue of settling up, when in fact they did not think about it: there is no agreement or trust to find or infer. Imputation corresponds to an exercise in normative analysis. An example of an imputation is deciding that the separating couple would very likely have agreed on a settlement based on joint wealth maximisation, with division of the ownership of the wealth set to support that maximisation and keep each party at least as well off as they might be under any other plan. Although it is easy to see how imputation in practice may be thought difficult to distinguish from the court’s deciding what it thinks would be ‘fair’ (the approach that the majority reject in Stack), it is different because of the possibility of adopting a coherent welfare standard as a guide. The House of Lords has been extremely wary of lower courts imputing common intentions to the parties, preferring to look for actual (express or implied) bargains typically creating beneficial interests under a trust. In Stack v Dowden, Lord Walker’s examination of Lord Diplock’s argument in Pettitt v Pettitt8 in favour of the court’s task being to impute the ‘common intention’ of the parties, using the whole course of the relationship as evidence, rather than relying more narrowly on evidence of an actual bargain, emphasises evidential difficulties rather than the disentanglement of inference from imputation. Lord Diplock acknowledged in Gissing v Gissing9 that his argument in Pettitt over imputation was not in the majority, but then uses the term ‘infer’ in a remarkably similar way to ‘impute’ in considering the same issues of beneficial interests, in fact joining in with the current muddle. Subsequent cases up to Stack v Dowden show the judges to be uncomfortable with the process of inference owing to a perception of unreality about the process and a recognition that it is often difficult in practice to maintain a distinction between inference and imputation.10 One could regard this reluctance as embodying a libertarian welfare principle, respecting the integrity

6

[2007] UKHL 17 (property division between an unmarried couple). ibid at [125]ff, which carefully discusses the distinction between inference and imputation that I make use of here. I am grateful to Jo Miles for discussions on the significance of the legal debate over the distinction between inference and imputation in relation to positivistic and normative economics. 8 [1970] AC 777 at 822E. 9 [1971] AC 886. 10 According to Griffiths LJ in Bernard v Josephs [1982] Ch 391 at 404, cited by Lord Walker in Stack v Dowden at [21]. 7

288 Antony W Dnes of the individual in choosing not to consider the issue of relationship breakdown rather than constructing the fix that the parties would have recognised if only they had been like some other, hypothetical, people. Having clarified the distinction between imputation (normative) and inference (positivistic), can we say anything about the current direction of the law on settling up between former cohabitants? I think so: one can in fact infer that the modern move toward equal division of joint property implies true imputation by the courts. Baroness Hale’s opinion in Stack,11 used by her to suggest a rebuttable presumption of equal shares, could be seen as a form of imputation: At first blush, the answer appears obvious. It should only be expected that joint transferees would have spelt out their beneficial interests when they intended them to be different from their legal interests. Otherwise, it should be assumed that equity follows the law and that the beneficial interests reflect the legal interests in the property.

Does this answer use the joint transferees’ failure to do anything else as inferred evidence, or move beyond factual matters to consider what rational people would have done had they thought about it? I interpret the resulting assumption as imputing rational assessment by the parties of their best interests, even where this may not have actually occurred: after all, doing nothing is in fact consistent with more than one outcome, such as its being too costly to spell out the beneficial interests, or the parties not having considered the matter. Silence is not generally acceptance, so acceptance is imputed here. To some considerable extent, the issue of imputation could be eased by providing a more solid foundation within which to work: an agreed welfare focus, in the terms favoured by economists. The normative approach coming out of mainstream law and economics would focus on situations where there is no evidence of a ‘bargain’, even allowing that the bargain could be inferred by citing an entire history of things said and done. The norm (‘arrangements likely to have led to joint-wealth maximisation’) would be deduced separately on the basis of what would have been in the joint interest of the couple, given that they had not agreed anything, in any terms, at the commencement of the relationship. The norm could then be widely debated and eventually emerge from legislation as a clarification of public policy. The major contribution of normative economics is in clarifying the steps taken in moving from setting policy objectives, that is, a clear target or set of compatible and weighted targets known as a ‘welfare function’ in the terms of economics, to checking how policies affect the target/s. Much debate in social policy areas such as those connected with cohabitation looks unfocused to an economist, in not clarifying objectives. Once objectives are set, and assuming recognition of the need for compatibility for multiple aims, the debate needs to move on to methods of achieving the objectives. Incompatibility between multiple objectives is not a particular objection to rational decision-making, but is a general objection to any solution to the underlying problem. 11

[2007] UKHL 17 at [54].

Rational Decision-making and Intimate Cohabitation 289 A good example of normative analysis, illustrating the need for greater clarity of reasoning, arises in comparisons of different standards that might guide settling up following divorce, and which might be extended to other forms of relationship break-up. Trebilcock (1994: 156) examines several arguments that all make a case for compensating the economically weaker party, typically, although not necessarily, the female, for the costs incurred as a result of the relationship.12 These costs often reflect the giving up of a career to stay home and raise children, which would be regarded, ex post, as ‘sunk costs’, or, equivalently, as ‘detrimental reliance’. Regardless of terminology, the key point is that the woman was initially committed to the relationship and gave up opportunities in anticipation that it would last and provide a range of benefits to her of both a pecuniary and nonpecuniary, and of both an individually and a jointly enjoyed, kind. Arguments over compensating the sunk costs associated with relationshipspecific investment by the female have generally settled into modern family law, but as just one factor to be recognised, actually in competition with others. The point is well recognised in recent English cases. In Miller; McFarlane,13 the House of Lords recognised three rationales standing behind judicial discretion exercised by virtue of the Matrimonial Causes Act 1973: (i) need; (ii) equal sharing of matrimonial property; and (iii) compensation. Much discussion in the literature is about possible conflicts between the compensatory principle and rationales (i) and (ii), as considered by Bridge (2006) in relation to the Law Commission’s (2006) consultation paper on the reform of cohabitation law. One can see that the notion of a rationale corresponds with an objective and that the three rationales taken together give a more complicated welfare function crying out pitifully for the application of a clear system of weighting. Every time the court applies the rationales it will be assigning weights, at least implicitly, to need versus sharing versus compensation. The exercise amounts to welfare economics writ small, more properly recognised as the court’s following what in the legal literature has for some time been referred to as a ‘liability rule’, that is, a rule whereby an objectively determined value is ascribed to the parties’ entitlements (Calabresi and Melamed, 1972). There are alternative normative approaches to settling up in the literature. Leaving aside the ‘I think it should be x per cent shares, and that’s that’ kind of argument, we can easily identify two serious alternatives. One is associated with Parkman (2002) who has argued for the use of a ‘property rule’, in the terms popularised by Calabresi and Melamed (1972), in which divorce is permitted by mutual consent only, and the court abandons its liability-rule, interventionist approach. A spouse seeking a divorce would compensate the other, who could insist on being at least as well off as before the divorce, as far as money can do 12 Trebilcock is critical of the reliance standard, arguing instead for a share in accumulated benefits that is likely to be kinder to women with low opportunity costs in entering the marriage. 13 [2006] UKHL 24 (assessing claims of equal shares in property and income in a longer and shorter marriage).

290 Antony W Dnes it, before consenting. The implied welfare standard of this approach is a variant on the requirement that gainers must be able to compensate losers and still gain from a change, before the change is permitted (Dnes, 2005: ch 1), often referred to as a Kaldor-Hicks test for a welfare improvement.14 Parkman’s approach at least preserves the expected benefits of the old marriage for both parties, because the divorce-seeker would back off if personal benefits from the change were not high enough to leave a surplus after compensating the other spouse’s losses. There is a strong sense here of finding the rule likely to be agreed on before the marriage, had the parties thought about a later possible divorce. Most folk do not want to consider such issues when getting married, so perhaps it behoves the court to come up with an approach reflecting the veto-entitlement rule that can be inferred or imputed in Parkman’s sense. Secondly, we could require the court to use a liability rule to award damages comparable to those emerging from Parkman’s bargaining approach. This variant would lead to the award of expectations damages to an unwilling divorcing party, since preserving at least the status quo benefits for him or her would be the minimum award that could be extracted by bargaining under the shadow of a veto entitlement. Compensation using expectation damages should normally exceed the reliance damages contemplated in Miller; McFarlane15 (Dnes, 1998). Preserving expectancy should stop ‘inefficient’ divorces, in which, as at present, the spouse seeking a divorce may be able to make off with personally enhanced benefits, regardless of the costs imposed on the other party. There are many more normative approaches that could be explored: for example, consider what settlement rule might be adopted under a Rawlsian veil of ignorance, which, in the context of family law, would involve not knowing whether one is male or female. Indeed, there is a significant struggle with the basics of a welfare comparison of compensatory approaches in the recent Law Commission (2007) recommendations concerning post-dissolution obligations between unmarried intimate cohabitants, and possible double-counting as a result of not quite getting to the bottom of the link between opportunities forgone and the generation of advantages flowing from cohabitation.16 The main point made here is simply a plea for clarity, in particular in distinguishing normative from the positivistic analysis to which we now turn. 14 The welfare standard is named after Lord Nicholas Kaldor and Sir John Hicks, who both popularised it in the 1940s, albeit in a form that accepted a hypothetical ability to compensate but did not require actual payment to be made. 15 [2006] UKHL 24. The House of Lords rejected expectations damages in Miller; McFarlane mainly in terms of avoiding a return to the earlier and abandoned ‘tailpiece’ to s 25 of the Matrimonial Causes Act that required the court to put the parties into the position they would have been in had the marriage continued: in actual fact an impossible thing to do and not strictly comparable to the expectations damages approach of contract law, which requires mitigation of costs by all and maintained expectancy for the breached-against party only (see Lord Nicholls’ dicta at [58]). The breaching party may be assumed to be better off if continuing to move on to greener pastures, notwithstanding a requirement to compensate his/her spouse. 16 This is the subject of a separate paper available from the author.

Rational Decision-making and Intimate Cohabitation 291 Positivistic Analysis Positivistic analysis typically uses insights from social theory, including information theory, economics and psychology, qualitative inquiry, and applied statistical analysis, in order to understand social changes concerning cohabitation. Examples would include the analysis of the response of marriage and divorce rates (Binner and Dnes, 2001), the age of first marriage (Mechoulan, 2006), and the age of first child-bearing (Brinig and Crafton, 1994) to legal changes such as the introduction of no-fault divorce, or no-fault settlements, and socio-economic changes such as increased female participation in the economy. Fisher and Low’s chapter 11 in this volume is a good example of statistical enquiry aimed at assessing the impact on spouses of the current divorce settling-up regime in the UK. The level of sophistication of these studies is high, increasingly eschewing data analysis at a point in time, which cannot capture secular change (that is, changes in attitudes and underlying institutions), in favour of pooling data over lengthy periods of time. Mechoulan’s (2006) paper is an excellent example of state-of-the-art approaches, based on panel data techniques, covering the statistical analysis of longitudinal data for substantial cohorts of people, and permitting variations over time and between the groups under study to be explored. Mechoulan links convergence of first-marriage rates across US states to increased marital selectivity by females in response to an increase in the risk and cost of divorce associated with the introduction of no-fault property settlements.17 It is fair to say that some clear results, for example the link between increased female labour market participation and marriage rates, have by now emerged in these studies, and would be very difficult to discern and explain in any other framework. The test that is generally applied to positivistic studies is to ask whether they predict some variable well: whether the theory and applied work give a compelling version of events. Judging whether theory and applied work give a compelling version of events is not a matter related to the unschooled judgement of the observer, although the choice to adopt a scientific methodology should still be regarded as a value judgement of sorts. Over the centuries, certain criteria have been recognised as encouraging fruitful research, such as preferring simple over complex theory if the explanatory power of the two is identical. More recently, attention to the lag structure of explanatory variables in statistics, for example checking whether an earlier shift in the explanatory variable is a significant influence on the dependent variable in the longer term, does give a basis for deducing causality between variables in certain well-defined cases (see the analysis of the growth of the US Army after the Civil War in relation to increasing Indian settler battles in Anderson and McChesney, 1994 for an interesting demonstration). In a positivistic study, in order to provide a satisfactory explanation of events, capable of predicting 17 The introduction of the no-fault divorce procedure, which has no significant impact in the study, should be distinguished.

292 Antony W Dnes the future course of those events, it is, strictly speaking, enough to show that the world is behaving as if a particular theory is correct. Quantification is usually important in that modelling, and it is likely to matter less what people think they are up to compared with independent indicators of their behaviour. A good example arises in family law in studies of marriage and cohabitation. In qualitative studies (see the discussion in chapter fourteen of this volume by Anne Barlow), interviewees often insist that money matters, or changes in family law affecting money, have nothing to do with their decisions to marry or cohabit. In a very narrow sense this may well be true, even leaving aside the issue of a natural tendency to answer such questions in a perceived tasteful manner. However, statistical studies, based more on hypothesis testing and relying on survey or census data, tend to show that such variables as the levels of divorce and marriage do respond to legal changes, and in particular to variables reflecting rules affecting financial settlements (Mechoulan, 2006) and to those reflecting welfare benefits and the workplace-based growing independence of women (Binner and Dnes, 2001). It is likely that many people would not have these financial influences in mind when answering a question framed in terms of simple legal or financial impetuses toward relationship structures. Another useful point of reflection is the introduction of legal aid for divorce cases in post-war Britain: this has been identified as a statistically significant influence on the level of divorce in our timeseries study (Binner and Dnes, 2001), but whilst legal aid clearly did make it easier for wives to initiate proceedings, it is unlikely that a questionnaire study would pick this up by asking about financial influences on the subject. The main purpose in separating normative and positivistic strands of analysis is to be clear as to the distinction between ought and is, but real world studies are often aimed at a mixture of complex questions. In practice, many studies will examine a variety of questions, some normative and some positivistic in nature, and one sometimes even encounters ‘normative analysis as positive theory’ (Viscusi, 2007) indicating that one predictor of behaviour could be the parties deciding to follow some aim on which they agree. In general, however, it clarifies debate if one understands when a prediction is being made about the effect of a legal change on the parties’ behaviour (positivistic), and the effect relative to a desired effect on their welfare (normative). I now consider some largely positivistic studies that focus on the feedback from legal rules on to human behaviour. Incentive structures, as noted by Fisher and Low (chapter eleven, this volume) are somewhat neglected in discussions of cohabitation legal reforms.

STUDIES OF THE NATURE OF MARRIAGE: LIFE PROFILES, OR ‘SHE GAVE HIM HER BEST YEARS’

The life profile theory of marriage is positivistic, and hence testable. It is associated with Cohen (1987, 2002), who then went on to make an influential normative

Rational Decision-making and Intimate Cohabitation 293 argument that, to avoid encouraging opportunistic divorce, the equivalent of contract (expectations) damages ought to be awarded to the fault-free party in a divorce. Actually, there is no reason to attach the liability to fault, as traditionally interpreted in family law, since for one spouse to indicate a wish for no-fault divorce is itself a breach of contract. The purpose of marriage, as a form of heavily-obligated cohabitation governed by a standard contract, could be seen as being to protect the economically weaker party from a form of exploitation taking the form of opportunistic behaviour emanating from an asymmetry in the life-cycle of men compared with women. This theory regards the defining aspects of marriage as being the rules governing divorce and post-dissolution obligations.18 According to Cohen, getting the rules of divorce settlement wrong, in his terms, risks creating an extremely worrying incentive structure characterised by opportunistic exploitation of the traditional wife. This concern is supported empirically by many papers, including Allen (1998), showing how the introduction of nofault divorce, far from simply releasing a one-time bottleneck of unhappy spouses, encouraged divorce, thereby adversely affecting the interests of traditional wives. Studies not supporting a permanent impact on divorce rates appear to be statistically faulty, or show adjustments occurring in marriage rates that are consistent with law having a permanent impact.19 Cohen’s argument is firmly based on a life-profile theory of traditional marriage, characterised by an asymmetry over the timing of investments made by the man and the woman as he specialises in the labour market and she specialises in homemaking. According to the life-profile theory, in a traditional marriage, the woman invests in children and home-making early on in marriage, incurring detrimental reliance by giving up alternative labour market opportunities, or, possibly, alternative marriage possibilities, and expects to remain with her husband enjoying the family income and home over the long term. He is freed of domestic responsibilities, enabling him to build up a career that will yield high earnings later on in the life-cycle, as is frequently observed (Bardasi and Taylor, 2008). Cohen

18 Note that in modern family law, the enforcement of obligations within marriage, eg for support, is becoming increasingly less important. Suits for spousal support are a rarity. Earlier work by Becker (1981) on the family was based on gains from specialisation in forming the family unit, which do not give a reason for marriage compared with simple cohabitation possibly with one’s granny. Modern work (Bardasi and Taylor, 2008) sees the gains from being married as at least ranging over those from specialisation, male human-capital formation (the enhancement of earning capacity discussed in Miller, McFarlane and in Charman v Charman (No 4) [2007] EWCA Civ 503), and signalling (of male sobriety, according to Akerlof, 1998). 19 Binner and Dnes (2001) used 50 years’ worth of UK data in a time-series model and showed that no-fault divorce laws had a permanent effect on divorce rates, and did not simply release a bottleneck. We noted that one of the frequently cited contrary (time-series) papers, Ellman and Lohr (1998), misinterpreted a spike in changes in divorce rates as release of a bottleneck, when a spike in changes is a long-term shift in levels. As noted earlier, the debate has moved on: Mechoulan (2006) finds a temporary effect from changing to no-fault settlement rules, in this case in a panel data study working on age of first marriage across US states and through time, but examines the response to the law, which includes increased search activity given unreliable marital promises.

294 Antony W Dnes observed that males more readily remarry in middle age than women,20 and might be tempted to abandon older spouses and move off without sharing the long-term returns of the marriage: the life profiles show an asymmetry between the sexes. In fact, her early investments have seriously reduced her ability to form a new relationship: while her child-bearing years may be over, she may still be encumbered by child-care responsibilities. Cohen’s analysis is economic in the broadest sense: the returns to marrying include companionship and related matters—the abandoning male may be looking for different company as well as retaining material assets, and the abandoned woman loses companionship as well as material claims. Its traditional focus needs updating to recognise a more egalitarian but still sexually unequal world, but is a starting point for analysis and is not a significant problem unless we think there are no asymmetries in life-cycles (on these persisting inequalities see further Scott and Dex, chapter three, this volume). The life-profile theory broadly implies that marriage may be a mechanism for protecting the female’s early marriage-specific investments over time in the face of possible later mistreatment by the male. Given uncertainties over the precise impact of domestic responsibilities, or growth in market-led family earnings and wealth, both male and female may wish to defer ultimately to court governance of their relationship by adopting a standardised marriage contract. He should wish to signal commitment to maintain a valuable level of support for her, and she would be encouraged to make the early marriage-specific investments. Marriages based on asymmetric relationship-specific investment patterns are particularly subject to possible opportunism after the introduction of no-fault rules that do not tie financial provision and property division to the equivalent of breach of contract, which is not necessarily fault as understood in family law. In terms of traditional marriage, it is now possible for the man to tire of an older wife and move on to a new relationship without maintaining the promised lifestyle, providing courts do not require the payment of expectations-based settlements. Divorce settlements based on meeting the needs or ‘reasonable requirements’ of an ex-wife, as in applications of the Matrimonial Causes Act 1973, especially before White v White,21 may effectively make divorce rather cheap, at least for wealthier men, and encourage such male opportunism (Dnes, 1998). The same is true of jurisdictions, such as Scotland or California, which require equal division of community property, since a half-share of the marital acquest may not yield the promised lifestyle. There is some empirical support for the worry that legal changes of this sort have increased divorce rates.22

20 Higher remarriage rates for divorced males appear as a statistical regularity in most international family law studies. And see Fisher and Low, ch 11, this volume on different repartnering rates by men and women post-divorce. 21 [2001] AC 596. 22 The modern trend to introduce reliance (opportunity cost) or restitution (returning the value of domestic contributions) will also under-compensate (certainly if taken as the sole measure of relief), as I explain in an earlier paper (Dnes, 1998). Think of the low opportunity cost of a waitress who marries a millionaire, or the low market value of hosting and domestic management compared with a promised wealthy lifestyle.

Rational Decision-making and Intimate Cohabitation 295 Lest we think it only the men who are capable of opportunistic exploitation, modern settlements could also make divorce too cheap from a female perspective: for example, the wife might tire of marriage and may know that, in a moderateasset setting, ‘needs’ could dominate contributions and lead to her keeping a significant proportion of the assets, particularly as a parent with care of minor children (see, for example, Hitchings, chapter nine, this volume). In terms of the debate over unmarried cohabitation, creating post-dissolution obligations could be a mechanism for protecting relationship-specific obligations, if there is some reason for thinking that there is an obstacle to marriage and the asymmetry over investment timing exists: this use of the life-profile theory and associated observations implies that we look for functional equivalence in designing marriage and cohabitation law.23 However, the failure of divorce law to enforce implied support promises, or to recognise fault, might well imply that unmarried couples of whatever kind could protect themselves better using private cohabitation agreements, although the complexity attached to these is potentially problematic.24 Ideally, we would not want a disaffected cohabitant to be able to move to a new relationship without having first to compensate the losses of the abandoned cohabitant, and it is doubtful that current marital law requires this of divorcing spouses. THE COASE THEOREM AND FAMILY LAW

There is a theoretical argument, following Coase (1960), associated with Peters (1986) and tested in many papers including Mechoulan (2006), that the legal regime affecting divorce settlement may not influence the level of divorce, or of marriage. A modern variant on the approach would suggest that requiring unmarried cohabitants to opt into, or opt out of, a standardised set of obligations would not affect the level of cohabitation, marriage, or single living. Such an argument depends upon bargaining to offset the effect of a legal change, and requires bargaining costs to be insignificant and the mechanism for offset to have no visible effect. A visible effect would occur, for example, if divorce rates settled down after an initial response but we discovered that fewer people were marrying. The analogy is with Coase’s argument that nuisance cases25 could, at least in a textbook sense, be settled by bargaining once the tortfeasor and victim knew their legal entitlements to carry out or stop the offending activity, resulting in

23 In Burns v Burns [1984] Ch 317 and in Marvin v Marvin 122 Cal Rptr 815 [App 1981] the unmarried ladies would be treated as if married. 24 There are good reasons for considering simplicity in legal doctrines as likely to enhance the quality of legal enforcement. See Dnes and Lueck (2009) for further details. 25 This note may help non-lawyers. Nuisance is the area of civil law settling conflict over resources (land-based activities). The court must decide whether conflicting uses of neighbouring land represent ‘reasonable’ uses of land, or not. The traditional remedy for nuisance is to stop (‘enjoin’) it whenever it is judged to be unreasonable, although financial compensation for continuing nuisance is developing in some jurisdictions. A nuisance roughly corresponds to an ‘externality’, or unpriced spillover effect, in economic analysis. Analysis of cases suggests that courts tend to see as unreasonable those uses of land that impose high costs on neighbours.

296 Antony W Dnes the best use of the underlying resources. In a case discussed by Coase, Sturges v Bridgman,26 a confectioner who could make more profitable, but noisy, use of property than a physician would carry on to the detriment of the physician’s medical practice, if the court refused to enjoin him. The value of the medical practice would be too low to enable the physician to buy out the noise. Conversely, if the physician received an injunction against the confectioner the physician could still be induced to allow the noise if the confectioner compensated him for lost medical fees. Either way, the same (joint profit-maximising) level of noise should result, as explained in more detail in Dnes (2005: ch 3). By analogy, if we were suddenly to move English family law from the current system of equitable distribution towards a community property system, or, say, move a communityproperty system from equal shares to 25 per cent or 75 per cent sharing, according to this theory it is possible that there would be no response in divorce, or marriage rates, since bargaining between the parties would reflect such changes. For there to be no response to the settlement rule, however, a number of heroic assumptions would need to be borne out in reality, or at least the world would need to behave as if they were true.27 These assumptions, apart from the obvious requirement for insignificant bargaining costs, include a requirement that all elements of value in the marriage could be rearranged, so that it would be broadly possible to redistribute early benefits from marrying to offset the change in postdivorce entitlements. Yet it may not be possible to divide all benefits from the marriage to facilitate bargaining: some assets/benefits may be inherently ‘indivisible’. Zelder (1993) examines such possibilities in detail and, importantly, notes that ‘indivisibilities’ really do exist that make bargaining difficult—children of the marriage whose future is to be determined by the court’s assessment of their best interests rather than efficient bargaining between the parties. The legal regime affecting settlement can allow marriages to end even when they could be reorganised to show a clear benefit to both parties, in which case, following the Coase (1960) theorem, one might have expected bargaining to take care of any problems. This inefficiency follows if one party can hold enough of the benefits of the marriage under the legal rules to put that amount beyond bargaining, which would otherwise focus on factors such as transferring beneficial use assets, consumption patterns, or altering the domestic division of labour. Such indivisibilities explain why there may be a failure to bargain in the shadow of the divorce court to rearrange individual benefits and prevent divorce occurring when there is a positive overall benefit from the marriage (Zelder, 1993; Brinig and Allen, 2000). In any case, the world does not behave as Coase’s theorem might suggest. The disruption of marriages by changes in divorce law is implied by empirical studies showing either a significant and permanent shift in divorce rates following 26 [1879] 11 Ch D 852 (injunction awarded to Sturges stopping a noise nuisance next to a Wimpole Street consulting room). Incidentally, the neighbouring properties are now occupied by law firms. 27 Note the positivism.

Rational Decision-making and Intimate Cohabitation 297 the introduction of no-fault divorce laws (Binner and Dnes, 2001), or increased search activity during marriage formation (Mechoulan, 2006)28 that affects marriage rates. Even if bargaining around legal change is a possibility, the world currently behaves as if bargaining is impeded. STUDIES OF COHABITATION

The move away from marriage and towards cohabitation represents one of the most significant shifts in society during the post-war period. The trends over 1960–2000 are similar across North America and many European countries, although detailed patterns may vary across countries.29 On average, first marriages fell from approximately 70 per 1,000 to 30 per 1,000 of the male population. The age at which first marriages occur has typically risen, with both men and women waiting an extra three years. Births outside marriage have increased from 5 per cent to over 40 per cent of all births. In addition, the proportion of cohabiting women between the ages of 20 and 50 has trebled.30 First, this change is clearly a major change in behaviour and is unlikely to be a random, inexplicable phenomenon. A puzzling aspect of the substitution is that cohabitation is against the interests of many women. Marriage is potentially a rather good mechanism for supporting long-term family investments, even taking a very approximate view of the fit of Cohen’s (2002) life-profile model, and without marriage women might predict their vulnerability to opportunistic behaviour (there might also be some men in such a position). It seems unlikely that changes in women’s economic activity and in techniques of child-rearing have reached a point where the sexes no longer show any asymmetric interdependence over life profiles.31 Therefore, one would expect a man’s willingness to offer marriage to remain a very important signal for young women, giving a ‘separating equilibrium’ distinguishing committed from uncommitted life partners. Inquiries have been made, but interviewed cohabitants give a variety of reasons for not marrying, or they believe erroneously that cohabitation is legally like marriage (Barlow, 2009), and, in general, one has an impression of tremendous confusion in the face of massive social change. Secondly, cohabitation has not generally been subjected to the same kind of settling-up regime as marriage in the event of dissolution. Up to now, there has been no intervention by a family court with powers to reallocate assets between partners or to create maintenance obligations.32 Cohabiting parties must rely

28 The argument that divorce rates may not change following legal reform because of bargaining adjustments is associated with Peters (1986). 29 See Kiernan (2002). 30 For a summary of recent demographic data, see Law Commission (2006: Part 2). 31 See Scott and Dex (ch 3), Price (ch 12) and Fisher and Low (ch 11), this volume. 32 But note the opinions in Stack v Dowden [2007] UKHL 17, followed in Fowler v Barron [2008] EWCA Civ 377, that support a rebuttable presumptive equal beneficial division of assets held jointly at law.

298 Antony W Dnes largely on natural ‘hostages’ that emerge in the relationship to limit opportunism. Such hostages may be provided by the presence of children, with whom parents may wish to maintain easy contact. Also, the search costs of finding a new partner may act to hold people together over a long period of time. In the case of marriage, such hostages will typically be bolstered by legal obligations between the ex-spouses to divide marital property and share earnings. Cohabiting parents have for some time been subject to child support obligations affecting all non-resident parents, regardless of marital status, to pay child support. Nonetheless, it is clear that it would be rational to choose to cohabit, if the parties actually wished to avoid, or perhaps in some jurisdictions just to lessen, the legal obligations towards each other in the event of dissolution. One can conclude from research findings that, to the extent that people think they are trying out partners (Lewis, Datta and Sarre, 1999) they are indeed avoiding marriage at that stage in their lives. ‘Trying out’ is consistent with the findings in Mechoulan (2006). From a life-profile perspective, if there were evidence that women were making early investments in family life that created benefits to be shared later, and that men were taking advantage of them by imposing cohabitation rather than marriage and then abandoning them later on, there would be the basis for a claim of exploitation. Even a tremendous libertarian might then reasonably seek to outlaw cohabitation by effectively turning it into marriage, or at least removing the basis for doing harm, just as we might wish to prohibit other frauds. Such reasoning might cause marriage-like obligations to be enforced when cohabitation came to approximate marriage (for example after children had been born within the union, or after a period of time during which one party had been economically dependent). However, the presence of exploitative cohabitation is not supported by the evidence (Lewis, Datta and Sarre, 1999) which suggests some form of shared expectations rather than one party creating an erroneous perception of common law marriage.33 At best one could claim that some people are misinformed about the likely outcome of cohabitation and the life-profile problem. There might then be a case for providing information and education, which the Ministry of Justice currently tries to do for England and Wales, but hardly one for the suggestion emerging around the world that would render consensual nonobligated cohabitation impossible without a specific agreement.34 A key issue, from a rational decision-making perspective, is that parties are free to avoid cohabitation as they could choose marriage. What if the meaningful sense of that freedom disappears? Choices are typically constrained. It may well be that 33 Common law marriage (ie an informal claim by a couple to be married) is recognised in some US states such as Montana. 34 The argument to change cohabitation obligations because people overestimate the obligations is perilously close to an argument that ‘the people could be different’: one of the fallacies of nirvana economics identified by Demsetz (1969), who argues that we need to fashion policy to reflect the welfare of people we actually have in society, not those whose preferences would be different if only they were wiser, fitter, younger, or whatever you wish.

Rational Decision-making and Intimate Cohabitation 299 the terms of cohabitation have changed in favour of males, following a reduction of pregnancy risk for unmarried females (Akerlof, Yellen and Katz, 1996), so that more women are willing to engage in unmarried sexual activity and fewer men seek marriage. Akerlof, Yellen and Katz (1996) demonstrate theoretically that there can be a tipping point in a population such that it moves from all single women benefiting from a male-provided marriage guarantee in case of pregnancy, whether they feel a need for the guarantee or not, to a position in which so many women no longer seek the guarantee that few couples end up married.35 There is no implication here that males are exploiting females, it is just that ‘terms of trade’ have moved adversely for women. Everyone is exercising free choice, under new constraints, but what to make of it? Indeed, the story told here is consistent with late twentieth-century developments, and with the reports on survey respondents’ views in qualitative inquiries (Barlow, chapter fourteen, this volume). Women are unsure why no one is marrying these days, whereas their mothers mostly did, and reassure themselves with the thought that it is all the same really. And, the men cannot believe they are getting away with so little commitment and also reassure themselves that cohabitation is really marriage, so as not to appear opportunistic. The alternative is to conclude that we are living among people who are systematically stupid or misinformed, which seems unlikely. Something really changed: after all, their parents probably36 had the same beliefs but did get married, as do people in jurisdictions with common law marriage. It is meddlesome, but entirely consistent with regulation in several areas, to intervene in an unregulated outcome such as the one just described if the outcome has wider implications in some way worrying for the rest of society. There are worries over the impact of relationship break-up on children, especially in relation to an absence of fathers (Popenhoe, 1996). There are also concerns that, once into unmarried cohabitation, the parties will begin to make joint pecuniary investments, for example in housing, and that hard-to-value non-pecuniary inputs may be involved (Bridge, 2006). So, in just the same way that certain free choices, such as those involving child labour or selling organs, are limited in broadly free and pluralistic societies (Novak, 1982: 55), so we might alter the legal constraints on cohabitation. It is important to realise that cohabitants are likely to respond to legal changes, and that there may be additional effects on the married and on the single. Rowthorn (2002) has argued that marriage has declined precisely because the courts have failed to keep the commitment signal clear: one can promise to love and support for life, and back out a few years later. Introducing a further form of 35 In some sectors of society the impact of the welfare system favouring single-person claims could also encourage such a result, as could greater female labour market participation. Akerlof, Yellen and Katz (1996) concentrate on the legalisation of abortion in a US context as the initiating change. 36 Although it is possible that previous generations did not have the same belief in common law marriage: see Probert (2008) for the argument that the common law marriage myth is of recent origin in England and Wales, and that common law marriage was not a part of the pre-eighteenth-century legal landscape in that jurisdiction.

300 Antony W Dnes marriage, which is what obligated intimate cohabitation would amount to, could further confuse the signals over commitment for couples. If people now think that unregulated cohabitation is marriage, what will they make of something like marriage but different? This reflection might also suggest caution over allowing enforceable pre-nuptial, or private cohabitation contracts: ultimately the clarity of signals of obligation and the quality of legal enforcement may depend on simplicity of doctrine. I have made similar observations in other areas of law, such as land-use servitudes (Dnes and Lueck, 2009). Also, there may be unintended consequences. The real importance of positivistic studies in family law and policy is in warning of what might happen if laws are changed. Is it possible that marriages might be displaced into less-obligated cohabitation? It is not good enough to say we do not think that will happen.

IN CONCLUSION: INFREQUENTLY ASKED QUESTIONS (IFAQS)

The practical results of my discussion of normative and positivistic studies of family law can be summarised in a number of answers to questions that should be more frequently asked than is typically the case. — Does economic rationality imply a financial motive for entering marriage or unmarried cohabitation? A. Not in general, as there could be a variety of reasons, or combinations of reasons, for marrying, all of which contribute to the welfare of the individuals concerned and affect their incentives. — Does this imply that legal and financial changes affecting dissolution obligations will have no effect on behaviour? A. No, not least because empirical work by economists, sociologists, and economist/lawyers picks up statistically significant impacts on marriage and divorce rates following on from legal change. — Should we just make marriage and cohabitation legally the same, given that so many people are receptive to the idea or believe in a common law marriage myth? A. The effect of this would either be to prohibit unobligated unmarried cohabitation in the absence of an agreement between the parties opting out of the scheme of financial remedies, or to create a new category of marriage—interpreted as a standardised set of obligations—and it is likely that this would have an effect on behaviour, possibly leading to undesirable repercussions. Normative questions would remain: in general, we have not allowed our legal approach to emerge from majority decisionmaking (consider capital punishment). — Are marriage and cohabitation really functionally equivalent? A. Apparently not, since empirical work of many kinds picks up differences between the two populations within a country and across countries. Social change has propelled a move towards cohabitation that differs in different places where social conditions differ, and that has increased across time where beliefs over

Rational Decision-making and Intimate Cohabitation 301 the legal status of cohabitants are unlikely to have varied, and has increased in places where common law marriage does exist, such as Montana. — Do studies of incentive effects from legal change suggest no need for legal support for cohabitants? A. No, there are issues concerning the current law’s encouragement of opportunistic behaviour. Most studies of incentive impacts suggest considerable caution over inadvertently making the cure worse than the disease by formulating policy that does not examine what amounts to regulatory impact. Most people seem to wish to avoid further destabilisation of relationships that include marriage and cohabitation. Before we consider whether a new rule should be adopted for either married or cohabiting couples, perhaps we should consider whether we are asking the right questions. BIBLIOGRAPHY Akerlof, GA (1998) ‘Men without children’ 108 The Economic Journal 287. Akerlof, GA, Yellen, JL and Katz, ML (1996) ‘An analysis of out of wedlock childbearing in the United States’ 111 Quarterly Journal of Economics 277. Allen, DW (1998) ‘No-fault divorce in Canada: its cause and effect’ 37 Journal of Economic Behavior and Organisation 129. Anderson, T and McChesney, F (1994) ‘Raid or trade? An economic model of IndianWhite relations’ 37 Journal of Law and Economics 39. Bardasi, E and Taylor, M (2008) ‘Marriage and Wages: A Test of the Specialization Hypothesis’ 75 Economica 569. Becker, G (1981) A Treatise on the Family (Chicago, IL, University of Chicago Press). Binner, J and Dnes, A (2001) ‘Marriage, Divorce and Legal Change: New Evidence from England and Wales’ 39 Economic Inquiry 298. Bridge, S (2006) ‘Money, Marriage and Cohabitation’ 36 Family Law 641. Brinig, M and Allen, DW (2000) ‘These boots are made for walking: why most divorce filers are women’ 2 American Law and Economics Review 126. Brinig, M and Crafton, S (1994) ‘Marriage and opportunism’ 23 Journal of Legal Studies 869. Calebresi, G and Melamed, AD (1972) ‘Property rules, liability rules, and inalienability: one view of the cathedral’ 85 Harvard Law Review 1089. Coase, RH (1960) ‘The problem of social cost’ 3 Journal of Law and Economics 1. Cohen, L (1987) Marriage, divorce and quasi rents: or, ‘I gave him the best years of my life’, 16 Journal of Legal Studies 267. —— (2002) ‘Marriage: the long-term contract’, in A Dnes and R Rowthorn (eds), The Law and Economics of Marriage and Divorce (Cambridge, Cambridge University Press). Demsetz, H (1969) ‘Information and efficiency: Another viewpoint’ 12 Journal of Law and Economics 1. Dnes, A (1998) ‘The Division of Marital Assets’ 25 Journal of Law and Society 336. —— (2005) The Economics of Law: Property, Contracts and Obligations (Boston MA, Cengage). Dnes, A and Lueck, D (2009) ‘Asymmetric Information and the Structure of Servitude Law’ 38 Journal of Legal Studies 89–120.

302 Antony W Dnes Ellman, IM and Lohr, SL (1998) ‘Dissolving the relationship between divorce laws and divorce rates’ 18 International Review of Law and Economics 341. Kiernan, K (2002) ‘Cohabitation in Western Europe: Trends, Issues and Implications’ in A Booth and A Crouter (eds), Just Living Together: Implications of Cohabitation on Families, Children and Social Policy (Mahwah, NJ, Lawrence Erlbaum Associates). Law Commission (2006) Cohabitation: The Financial Consequences of Relationship Breakdown, Law Com CP 179 (London, TSO). —— (2007) Cohabitation: The Financial Consequences of Relationship Breakdown, Law Com No 307 (London, TSO). Layard, R (2005) Happiness (London, Allen Lane). Lewis, J, Datta, J and Sarre, S (1999) Individualism and Commitment in Marriage and Cohabitation, Research Paper 8/99 (London, Lord Chancellor’s Department). Lipsey, RG (2008) ‘Positive Economics’ The New Palgrave Dictionary of Economics, 2nd edn. Mechoulan, S (2006) ‘Divorce Laws and the Structure of the American Family’ 35 Journal of Legal Studies 143. Novak, M (1982) The Spirit of Democratic Capitalism (New York, Simon and Schuster). Nozick, R (1974) Anarchy, State, and Utopia (New York, Basic Books). Parkman, A (2002) ‘Mutual consent divorce’ in A Dnes and R Rowthorn (eds), The Law and Economics of Marriage and Divorce (Cambridge, Cambridge University Press). Peters, HE (1986) ‘Marriage and divorce: informational constraints and private contracting’ 76 American Economic Review 437. Popenhoe, D (1996) Life Without Father (Cambridge, MA, Harvard University Press). Posner, R (1984) The Economics of Justice (Cambridge, MA, Harvard University Press). Probert, R (2008) ‘Common-law marriage: myths and misunderstandings’ 20 Child and Family Law Quarterly 1. Rawls, J (2005) A Theory of Justice (Cambridge, MA, Harvard University Press). Rowthorn, R (2002) ‘Marriage as signal’ in A Dnes and R Rowthorn (eds), The Law and Economics of Marriage and Divorce (Cambridge, Cambridge University Press). Trebilcock, M (1994) The Limits of Freedom of Contract (Cambridge, MA, Harvard University Press). Viscusi, WK, Vernon, JM and Harrington, JE (2007) The Economics of Regulation and Antitrust, 4th edn (Cambridge, MA, MIT Press). Zelder, M (1993) ‘Inefficient dissolutions as a consequence of public goods: the case of nofault divorce’ 22 Journal of Legal Studies 503.

14 Legal Rationality and Family Property What has Love got to do with it? ANNE BARLOW

INTRODUCTION

F

amily law, in so far as it is trying to regulate disputes relating to money and property on relationship breakdown, has its work cut out. For it is trying to deal in a rational way with issues between people who are in an emotionally charged relationship situation and where most might predict that rationality is unlikely to prevail (Beck and Beck-Gernsheim, 1995). Whilst some, such as John Dewar, see the answer as conceding to what he terms ‘the normal chaos of family law’ which nonetheless works on a practical level due to the pragmatic solutions of professionals operating within a discretionary legal framework (Dewar, 1998), others have argued that this is not the optimal way forward for family law in general (Henaghan, 2008) and the regulation of new family forms in particular (Barlow et al, 2005). This chapter will therefore consider whether family law can avoid the trap of a ‘rationality mistake’—whereby legislators overestimate the law’s ability to steer behaviour in a particular direction (Barlow and Duncan, 2000; Barlow et al, 2005)—yet still develop a coherent theory of family law to apply in this field (Eekelaar, 2006; Henaghan, 2008). It has been argued convincingly by critical theorists that family law ‘needs to be socially located’ (see eg Freeman, 1985: 153–54). Given shifting attitudes and more complex married and unmarried families resulting from changed parenting, partnering and repartnering patterns and behaviours, this presents a real challenge. In rising to this, it will be argued that it is now time to take stock of both the emotional and economic foundations and commitment on which modern couple relationships are built in order to consider how family law should weigh the competing values of promoting personal financial autonomy yet providing legal protection for the economically weaker partner on relationship breakdown. Arguably this has already been done in the

304 Anne Barlow cohabitation context by the Law Commission in its consideration of proposals for the reform of cohabitation law (Law Commission, 2007). But has the right balance been struck here? Is the current legal hierarchy still fit for purpose or are we drawing the regulatory lines in the wrong places? These are the questions this chapter aims to pursue. In so doing, it will draw on empirical research to examine whether family law in this area can find a way to cope with its chaotic raw material, avoid the legal rationality mistake yet become sufficiently coherent to provide satisfactory outcomes for those it serves. LEGAL RATIONALITY AND THE NORMAL CHAOS OF LOVE

At its foundation, law is a system predicated on its power to bring about what might be termed ‘legally rational’ behaviour. It is a closed system which assumes that its deterrents and rewards are known (or should be known) and will shape behaviour. Whilst this classical positivist formulation of the working of the legal system has been much critiqued (see, eg, Dworkin, 1977), an assumption of ‘legal rationality’ still underlies the way law is expected to work. This approach remains visible in the family law field, where different styles of family receive different legal treatment in different contexts but according to a preferential hierarchy which has developed in an ad hoc manner, with only sporadic legislative review. The assumption of a legally rational reaction to family law is certainly a matter which has influenced legislators and which dominates discussions surrounding any family legislation which might potentially detract from more traditional moral standards. Whilst the same-sex civil partnership legislation (the Civil Partnership Act 2004) was able neatly to side-step allegations of undermining heterosexual marriage (with which, after all, it was not a direct competitor), by focusing debate on the government’s equality and non-discrimination agenda (Department of Trade and Industry, Women and Equality Unit, 2003), attempts to introduce greater legal rights for informal cohabitants—often dubbed ‘marriagelite’ (Morgan, 2000)—are seen to pose a greater threat. This has been observed most recently in the government’s cautious reaction to the Law Commission’s recommendations for reform of cohabitation law (Law Commission, 2007). Rather than acknowledging the case for reform made out very strongly in the Law Commission’s report, more research into Scotland’s experience of similar but very recent legislation was deemed necessary prior to legislative action.1 Despite the already overwhelming statistics showing the decline in marriage and increase in heterosexual cohabitation (National Statistics Online, 2008), the belief in law’s power to influence the future direction of these social phenomena remains undiminished and is deeply influential within party politics. As Baroness 1 Ministerial Statement to Parliament by Bridget Prentice, Parliamentary Under-Secretary of State for Justice, Hansard HC vol 472 col 22WS (6 March 2008).

Legal Rationality and Family Property 305 Young asserted in the parliamentary debates surrounding the Family Law Bill 1996 proposing divorce reform, ‘law influences behaviour and it sends out a very clear message. There would be no point in legislating at all if law did not influence behaviour.’2 It is this concept of legal rationality which is used in this chapter where its role as a useful tool or inhibitor of progress in the family law field will be explored. Love, on the other hand is the very antithesis of rational thought and behaviour, legal or otherwise. Indeed, sociologists such as Beck and Beck-Gernsheim (1995) talk convincingly of ‘the normal chaos of love’, a phrase which encapsulates the way in which rationality is overpowered by love, which is something far stronger and beyond control, leaving the ‘victim’s’ reason in a distant parallel universe and at least temporarily beyond recall. When the first flush of love fades, a semblance of rationality may return, but its roots are not deep. It can easily be dislodged by the all-too-powerful emotions which accompany love’s rejection or a fear that love might be lost. Thus when relationships based on love break down, rationality is not restored. On the contrary, often more chaos is fuelled, which poses the question of law’s suitability for resolving disputes born out of such emotionally charged situations. This leads us to ask whether family property disputes are social problems for which law, at least in its present form, is altogether unsuited or whether it is indeed possible to use law to help find more imaginative solutions. Dewar (1998) identified the normal chaos of family law, endorsing Bourdieu’s suggestion that the logic of following a rule ceases at the point at which logic ceases to be practical. How practical are legal rules in a sphere dominated by emotions where diverse social (and behavioural) norms are no longer reined in by a set of moral and religious values shared across society that can be reflected in law? Should we instead be asking ourselves how long can we go on pretending that society is well served by such a mismatching of social problems and legal solutions? Or is this to overstate the case? Can people, even those in love-based relationships, be persuaded to take legally rational steps to protect their interests? According to Barlow et al (2005), to assume this is generally possible is to fall prey to the legal rationality mistake, whereby law assumes people act according to the logic of the law whereas in reality they act in accordance with the social imperatives within their own lives: [P]olicymakers risk falling into what we have called a ‘rationality mistake’ … They may create policy assuming a particular sort of public behaviour and rationality, whereas in fact most people make their decisions according to different criteria … according to moral, relational and emotional judgements of what is the proper thing to do in their situation. To preserve this rationality mistake by not reforming cohabitation law, may be to risk a permanent rupture between social and legal norms. This could completely undermine the credibility of the law in relation to families.’ (Barlow et al, 2005: 97–98)

2

Hansard HL vol 569 col 1638 (29 February 1996).

306 Anne Barlow Avoiding the Legal Rationality Mistake How then might law avoid the rationality mistake? Barlow et al’s view was formed following a two-year study focused on the social and legal attitudes of a nationally representative sample of people in England and Wales. One key finding was that in the British Social Attitudes Survey conducted in 2000, the majority of people (56 per cent) and a bigger majority of cohabitants (59 per cent) believed in a ‘common law marriage myth’ whereby people falsely believed cohabiting couples have a ‘common law marriage’ in which they gain the same rights as married couples after a period of time (Barlow et al, 2001, 2005). Another important finding, however, was that the vast majority of those cohabitants who did not believe this myth and were aware of the legal position they were in did not respond by taking available legal steps such as making wills or declarations of ownership regarding the family home, often despite a desire and intention to do so. Reasons for this included the cost of legal advice, the perceived need for advice only when things go wrong, the complexity of the various legal steps needed, and a general inertia surrounding putting one’s affairs in order (Barlow et al, 2005: 78–79). One possible solution if one does not want to reform the law is to make people more legally aware and to better facilitate the taking of appropriate legal steps; and this was indeed the government response. A web-based public information campaign, The Living Together Campaign (www.advicenow.org.uk/living-together/), which aimed to inform cohabitants of their legal situation and help dispel the common law marriage myth, as well as provide downloadable forms to enable people to take the requisite legal steps more easily and cheaply, was funded by the Department for Constitutional Affairs (now the Ministry of Justice) over a threeyear period from summer 2004. However, this campaign was launched against the background of a legal culture which, it could be argued, has actually discouraged those in love from being legally rational about family property. Its unwillingness to enforce pre-marital agreements on divorce3 in a sense endorses the view that legal rationality and love do not and should not go hand in hand. Similarly, the Law Commission’s strong recommendation that, regardless of whether other reform was enacted, cohabitation agreements should be made unequivocally enforceable and explicit confirmation that they are not regarded as contrary to public policy be provided (Law Commission, 2007: [5.8]–[5.9]) has so far gone unheeded by Parliament. Add to this the inconsistencies within our law as to whether to treat cohabitants as married—as in the case of social security, tax credits, and private tenancy succession law; two single individuals—as in the case of capital taxes and pensions; or somewhere in between—as for compensation under the Fatal Accidents Act 1976 or a claim from a deceased partner’s estate under the Inheritance (Provision for Family and Dependants) Act 1975), and one begins to understand why the falsely 3 F v F (Ancillary Relief: Substantial Assets) [1995] 2 FLR 45; NG v KR [2008] EWHC 1532; cf K v K (Ancillary Relief: Pre-nuptial Agreement) [2003] 1 FLR 120.

Legal Rationality and Family Property 307 supposed existence of common law marriage (Barlow et al, 2005, 2008a) seems far more credible than the actual state of the law and the legally rational yet disparate and complex steps which it consequently requires. If public information might not then be the whole answer, what else might be considered? Could family law mould itself differently so as to better fit the needs of the community it serves? Is there any public consensus on how law should tackle these problems which can be used to better socially locate the role of family law as the arbiter in the (gendered) power struggle which often erupts on heterosexual relationship breakdown? Is there, perhaps, an optimum moment at which to focus a couple’s minds on a marital or cohabitation agreement perhaps, if private ordering is part of the solution? What about views of those who have experienced marriage or cohabitation on what the law should do for people when relationships break down and there are disputes about property and postrelationship support? Is there any room for a bottom-up approach to legislating in this sphere, where the very nature of family law’s impact on a person’s exercise of their right to private and family life (guaranteed by Article 8 of the European Convention on Human Rights) might justify the legislature taking into account public attitudes in this sphere more than in any other? In other words, how best can we deliver family law solutions in this field which are socially located? In order to address these issues, let us now consider the messages from recent research as it relates to relationships, separation, property, and attitudes to legal reform in this field.

MARRIAGE, COHABITATION AND FAMILY PROPERTY—TIME FOR RESPONSIVE MODE?

Research confirms that the decision whether to marry or cohabit is now considered a lifestyle choice by the vast majority of people in Britain, and the legal consequences of these different family forms have little or no bearing on the choice of family structure, at least on entering a first relationship (Barlow et al, 2005; Hibbs, Barton and Beswisk, 2001; cf Dnes, chapter thirteen, this volume). Love clearly triumphs over legal rationality in this situation. Similarly, the need for ‘a proper wedding’—a powerful social symbol—trumps any desire to have the security of a legal contract in the eyes of many. A lawyer’s response to this might well be ‘caveat emptor’—let the buyer beware—and if you know you are not married, then you know you are getting something other than the legal contract of marriage, however marriage-like you might think your relationship is. However, if despite the government-funded information campaign people still wrongly believe that cohabiting for a period of time gives couples the same legal rights as if they were formally married, as the most recent British Social Attitudes Survey suggests (51 per cent of people in general and 53 per cent of cohabitants so believe (Barlow et al, 2008a)), then this gives pause for thought. Furthermore, the proportion of people who correctly do

308 Anne Barlow not believe in the myth has remained constant (37 per cent) since 2000 (although among cohabitants this had increased from 35 to 39 per cent), but there has been a growth in the percentage of those who are unsure of the legal position which has risen from six to 10 per cent. Whilst there was a significant increase in the numbers who agreed that marriage provided greater financial security than cohabitation, this still left less than half of all cohabitants agreeing with this view, which is a matter of concern given the financial consequences of the legal disparities between the two styles of relationship. The Living Together Campaign, whilst moving things in the right direction, has not, according to these results from the latest British Social Attitudes Survey, significantly dismantled the common law marriage myth. Neither has it persuaded large numbers of people to put their cohabitation affairs in order, either by marrying or by taking appropriate legal action, such as making wills (only 12 per cent of cohabitants in the recent national study had done so), taking legal advice (19 per cent had done this), or entering into cohabitation agreements. There is some evidence of an increase in the number of declarations of trust relating the family home made when property is jointly purchased (15 per cent, compared with eight per cent in 2000), probably in large part due to the change in Land Registry Rules requiring declarations to be lodged prior to registration of a purchase. The study of the campaign’s effectiveness on legally aware cohabitants (Barlow, Burgoyne and Smithson, 2007) showed how difficult it is to have any great impact even with a well-publicised media and website campaign that provided free legal information and documentation and was relatively easy to use. Although there were some positive outcomes, particularly on how useful the site was perceived to be, most people found it too difficult to take action for a variety of reasons. Even when people wanted to behave in a legally rational way, there were many obstacles. Reasons given included: there was no appropriate action that could be taken to achieve what the couple wanted; a fear that action might cause problems in their relationship; that a partner would not agree to the proposed action; or, for the largest category, they had not yet got around to it. Others mentioned the cost, either financially or emotionally, of legal action against the likelihood of needing the benefits of that action at a later stage. There was also a recurring view that filling out a living together or cohabitation agreement was ‘too negative’. Optimism also made it difficult in a relationship which was going well to think of things turning sour: You’ve got to weigh up the cost of these sort of legal transactions and think, well, …, is it really worth going into it for, you know, something that’s probably not going to happen. (Laura, cohabitant) Overall, it was concluded that policies directed at encouraging legally rational behaviour over relationship choice were useful but not transformative in changing behaviour, impacting only at the margins (Barlow et al, 2007, 2008). Thus both the continued existence of the common law marriage myth and the lack of behaviour transformation through awareness campaigns pose problems

Legal Rationality and Family Property 309 for the continuation of family law’s current hierarchy of relationships where law privileges marriage and now civil partnerships, yet deals with the increasing numbers of informal relationships differently in different contexts. Another more radical response might be for law to give people the rights they think they already have. For if people are making decisions on a false premise but which they sincerely believe is imbued with a ‘common law’ legal truth drawn from the lived ‘social truth’ that married and cohabiting couples are functionally similar and deserving of equal treatment, particularly where there are children, should the law then not consider aligning itself with that social truth? This is the approach that has been taken in other common law jurisdictions such as Australia and New Zealand (Atkin, 2001, 2003; Graycar and Millbank, 2007) but is met here with the criticism that to go down this road is oppressive to those who are deliberately avoiding marriage and its legal obligations and encourages women’s dependency on men (Deech, 1980; Freeman, 1984). Certainly research has characterised different types of cohabitant. Smart and Stevens (2000) identified a continuum of cohabiting and married relationships which range from the contingently committed to the mutually committed, with more marriages falling at the mutually committed end and cohabiting relationships being far more mixed. Lewis’s study revealed most cohabitants in her sample as mutually committed rather than individualistic (Lewis, 2001), and Maclean and Eekelaar’s study (2004) did not find that the moral obligations felt towards family members in unmarried as opposed to married families differed significantly. Barlow et al (2007) found diversity among their sample of legally aware cohabiting couples who had used the Living Together Campaign website. This included a typology of four main styles of couple—ideological, pragmatic, romantic, and uneven—and concluded that different legal solutions were needed to meet the different needs of each group. Protection of uneven cohabitants, where one of a couple wanted to marry or take legal action whilst the other did not, was seen as particularly necessary, for these couples could never agree to resolve the situation for themselves, leaving one partner very vulnerable and in need of legal protection, particularly on relationship breakdown. Enforceable cohabitation agreements which were accompanied by an acknowledged legal status similar to but other than (as they viewed it) patriarchal marriage would, on the other hand, be very welcome to ideological couples. Pragmatic couples would all be prepared to take legal steps or even marry for legal reasons if necessary, even though they may delay taking such action and would prefer not to have to do so. Romantic couples, on the other hand, many of whom were in a trial marriage and were building towards a big wedding that might still be years away, felt it would be wrong to marry for purely legal or financial reasons. However, very few (10 per cent) even in this study of legally aware cohabitants, many of whom were ideologically opposed to marriage, were opposed to the extension of legal rights currently enjoyed only by married couples to cohabitants with children. Nor were views expressed that this type of legislation would be oppressive, providing a couple could jointly agree to opt out from the presumptive scheme.

310 Anne Barlow At the time when the Law Commission was looking at reform of cohabitation law on separation and death, social attitudes nationally were again tested in the British Social Attitudes Survey 2006 and follow-up study (Barlow et al, 2008a, 2008b). This looked, inter alia, at which behaviours within different styles of (heterosexual) cohabiting relationships warranted claims for financial support on relationship breakdown or death. As with the Living Together Campaign Study discussed above, there were few signs that an extension of legal rights would be viewed as oppressive and this was particularly the case where there were children of the relationship. Public attitudes were tested by putting forward 10 scenarios which explored how meritorious a partner’s claim for financial provision was viewed where the law currently metes out different treatment to married and cohabiting couples on relationship breakdown on death and which was under review by the Law Commission. The aim was to see how views changed according to variables such as: the presence or absence of children; whether the parties were married; the length of the relationship; financial and domestic contributions to the relationship; and the circumstances leading to the claim for financial provision. Table 14.1 below summarises the findings of what was viewed as appropriate on relationship breakdown. Table 14.1: Beliefs regarding Rights to Financial Provision on Separation for Married and Unmarried Couples % agree partner should have right to financial provision on separation if …

If couple not married

If couple married

… couple living together for 20 years, 3 children, woman reduced work to part-time and then gave up work to look after family and home, man supported family financially and owns home, woman has no income and poor job prospects.

89

N/A

… couple for ten years, no children, one partner worked unpaid to build up other partner’s business, partner who runs business also owns family home, other partner has no property or income of own.

87

93

… couple for ten years, one partner has well-paid job requiring frequent moves, other partner has worked where possible but has not had a settled career.

69

81

… couple for two years, one has a much higher income than the other and owns the family home.

38

62

… couple living together for two years with young child and now separating. She will be child’s main carer and he will pay child support.

74

N/A

3197

3197

Base

Legal Rationality and Family Property 311 This confirmed a different hierarchy of relationship status after short couple relationships to that embodied in law. Similarly, when asked about the situation where a cohabitant died, even after a two-year short childless relationship, there was clear support for a functional approach to regulating couple relationships. Two-thirds felt the same remedies should be available to cohabitants as married couples, with 98 per cent believing this after a 10-year childless relationship. This research found little support nationally for the law to distinguish between financial remedies for separating married and cohabiting couples where the relationship was long term; where there are children of relationship or where there is evidence of joint enterprise contributions, for example to a business. There was evidence that views change across the generations and this supports the idea put forward by Eekelaar (2006) that it might be important to bear this in mind when attempting to legislate for future generations in this field. Whilst it is difficult to know for certain if this variance in views is a periodic or generational effect which may change as people get older, younger people without children tended more readily to endorse the promotion of financial autonomy on relationship breakdown. However, this was more so in scenarios which did not involve children. Most people were found to be very child-centred in their thinking, which does reflect one of the main tenets of family law. Those who were older or who had more than one child tended to favour more equal recognition of financial and non-financial contributions, and where there was a child in the scenario this approach broadened across all categories of research participants, endorsing the functional approach favoured by participants in earlier research (Barlow et al, 2005; Cooke, Barlow and Callus, 2006; Barlow, Burgoyne and Smithson, 2007).

Why does this Differ from what Law Currently Does? If law could justify its different treatment of approaches to family property in the married and cohabitation contexts in a convincing manner, then perhaps the legally rational approach could be endorsed. As things stand, whereas cohabitants’ assets are divided according to property law on relationship breakdown with little or no attention being given to the nature of the relationship (Burns v Burns;4 Stack v Dowden5), a completely different approach is taken on marriage breakdown where the aim is to achieve fairness as between the parties. For financial and non-financial contributions to the marriage are viewed as equal and justify an equal division of assets on divorce, unless other factors such as the needs of the parties, a truly exceptional contribution by one party alone, or inherited wealth or other non-matrimonial assets justify a different approach 4 5

[1984] Ch 317. [2007] UKHL 17.

312 Anne Barlow (White v White;6 Miller v Miller; McFarlane v McFarlane7). However, much of the logic as to why we do as we do in the married context, has been lost. Why is an economically weaker spouse rewarded for the public commitment given on marriage at the point of divorce when that commitment has been broken? We have no-fault divorce and so assets are in the vast majority of cases redistributed regardless of who is to blame—we are not compensating an innocent party for a breach of contract, as was reiterated forcefully by the House of Lords recently in Miller; McFarlane.8 Often, quite the reverse is true. On divorce, pre-marital cohabitation may in effect lengthen the duration of the marriage for the purpose of deciding how assets should be redistributed, accepting that the private commitment of cohabitation is valuable if it translates into marriage.9 Yet we do not redistribute assets on the basis of a long-lasting private commitment if a long cohabitation relationship breaks down. If the rationale behind the messages being sent by the current state of the law is itself confused and at odds with what is socially acceptable, clear messages are not likely to get through to the public, let alone encourage ‘legally rational’ behaviour. Either we need to accept this as Dewar’s normal chaos of family law or questions need to be posed about what law is doing and whether it can justify the different treatment of cohabitants from that of functionally similar married couples. What are we rewarding on divorce exactly? If it is the fact that one party has sacrificed their own financial position for the greater good of the family, perhaps by giving up work to care for children, why is this rewarded only in the married context when exactly the same benefits are being gained by the partner, children and, indeed, society when cohabitants order their lives in this way? The empirical evidence clearly shows that it is not public attitudes nor, indeed, those of cohabitants that are standing in the way of such a development. Furthermore, the strength of public support for maintenance on separation following a short childless relationship is also questioned by the results from the most recent British Social Attitudes Survey (see Table 14.1 above). Whilst there was little support for cohabitants without children who separate after two years to receive financial provision (just 38 per cent were in agreement), more surprisingly there was less support for maintenance following a two-year childless marriage, where 62 per cent agreed this was appropriate, than for maintenance for the primary caring cohabitant following a two-year cohabitation relationship with one child, for which 74 per cent were in favour. This begs the question of whether the regulatory lines have been drawn correctly in providing financial provision on separation. Clearly to use marriage

6

[2001] 1 AC 596. [2006] UKHL 24. 8 ibid. 9 See, eg W v W (Judicial Separation: Ancillary Relief) [1995] 2 FLR 259; GW v RW [2003] EWHC 611. 7

Legal Rationality and Family Property 313 and now civil partnership as the sole determining factor for providing financial provision is convenient from the point of view of the legal system. Whether or not you hold the required status is easy to prove by production of the marriage certificate or civil partnership document. There may be more arguments about whether or not a couple fall within the definition of cohabitants, yet this is something that the courts already deal with quite competently in those areas where rights have been extended to such couples. Ease of regulation cannot really stand up as a very good reason not to endorse a change of approach, particularly where this has been achieved in other common law jurisdictions with seemingly minimal difficulty. The feminist argument that the law should, at the beginning of the twenty-first century, be discouraging rather than promoting further financial dependence of women on men might hold more sway. Yet where there is relationship-generated need or disadvantage which falls principally on women, who are most often the primary carers of children in all styles of relationship, should the law continue to discriminate against those who have done this outside marriage yet not require financial autonomy from divorcing childless women in the name of fairness? The last and most cited justification for the continuance of the status quo is so as not to undermine marriage. Marriage is more stable than cohabitation and is seen as special. Indeed, only nine per cent of the public in the most recent British Social Attitudes Survey agreed that marriage is only a piece of paper (Barlow et al, 2008). Yet, retaining the status quo has not in this jurisdiction stemmed the tide of those cohabiting nor, indeed, the decline of the numbers marrying each year. Policy documents from both the major political parties talk of supporting marriage in the name of creating more stable families for children (Home Office, 1998; Social Justice Policy Group, 2007). However, there is an argument that it is those families that are likely to break up that need most to be regulated. Certainly, research shows that the current law can produce great economic disadvantage for cohabitant parents on relationship breakdown (Tennant, Taylor and Lewis, 2006; Lewis, Tennant and Taylor, chapter eight, this volume). Furthermore, where is the logic in encouraging marriage whilst providing a perverse incentive not to marry by denying legal remedies for other styles of families with children who are in need of them when relationships break down? Generally, it is not clear how extending rights and obligations which appertain to marriage to functionally similar cohabiting couples will in any way undermine marriage further than is already happening. What is it that we are concerned about in extending remedies to the cohabitation context, given the seemingly uncomplicated experiences of other jurisdictions that have awarded legal status and family property rights to cohabitants (see eg Atkin, 2001; Graycar and Millbank, 2007) and the fact that countries that have gone down this route have not suffered any observable decline in marriage rates above and beyond existing rates of decline (Kiernan, Barlow and Merlo, 2007)?

314 Anne Barlow RECONSIDERATION OF HIERARCHY TO REFLECT CHANGING CIRCUMSTANCES?

The Law Commission for England and Wales has recently reviewed the law governing the financial consequences of cohabitation breakdown and, after a twoyear project in which it consulted widely, it made recommendations for reform of cohabitation law in its report in 2007 (Law Commission, 2006, 2007). However, the idea that cohabitation law should be put on a par with marriage was rejected. In its consultation paper, the Law Commission explained the view as put to it by one consultee: Some account should be taken of the decision of the parties not to marry: it is one thing to relieve the unequal impact of the relationship, but quite another to treat the parties as if they had actually married. (Law Commission, 2006: [6.239])

The report confirmed this view: Applying the MCA would impose an equivalence with marriage which many people would find inappropriate, and some consultees suggested that it is unlikely that a scheme which equated cohabitation with marriage in this way would be politically attainable. (Law Commission, 2007: [4.8])

Rather it put forward a radical new presumptive scheme which does not mirror marriage but which proposed acceptance of cohabitation contracts and redress for relationship-generated economic disadvantage or retained benefit. The Law Commission recommended that its scheme should apply to all cohabitants with children and other childless cohabitants (whether of same or different sex) who have lived together for a qualifying period, which they suggested should be between two and five years. Couples should be able to opt out of the scheme in order to protect individual autonomy providing certain conditions were satisfied, notably that the agreement was in writing, signed by both parties and made clear the parties’ intention to disapply the scheme (Law Commission, 2007: [5.56]). In the case of manifest unfairness, a court would have power to set aside an opt-out agreement ([5.61]). The idea of the scheme was to base a former cohabitant’s claim upon the ‘economic impact’ of cohabitation. On separation, an eligible applicant in making a claim against their partner would have to prove that either the respondent had a retained benefit, or the applicant had an economic disadvantage as a result of qualifying contributions the applicant had made. It recommended that the same style of orders would be available for financial provision as is currently the case on divorce or civil partnership dissolution with the exception of periodical maintenance payments, although the grounds for making the orders were to be quite different. A qualifying contribution was defined as any contribution to the shared lives or welfare of members of the family, and could be non-financial ([4.34]). An economic disadvantage was stated to be a present or future loss. It could include a diminution in current savings as a result of expenditure or of earnings lost during the relationship, lost future earnings, or the future cost of paid child care ([4.36]).

Legal Rationality and Family Property 315 A retained benefit, on the other hand, could take the form of capital income or earning capacity that had been acquired, retained or enhanced ([4.35]). The remedy in respect of retained benefit was straightforward. The court would be able to order the reversal of any retained benefit ‘in so far as it is reasonable and practical and having regard to the discretionary factors’, which were: — the welfare of any minor child of both parties (which was to be the court’s first consideration); — the financial needs and obligations of both parties; — the extent and nature of any financial resources of each party now and in the foreseeable future; — the welfare of any child living with, or who might reasonably be expected to live with, either party; — the conduct of each party, defined restrictively but including a qualifying contribution made despite the express disagreement of the other party. With regard to a finding of economic disadvantage, however, how exactly this would be calculated in practice was cause for concern among commentators, with a crystal ball being seen as a useful piece of equipment in the calculation of lost future earnings (Probert, 2006). In any event, only one half of any economic disadvantage would be paid over according to the principle that any loss should be shared equally, and in making any order, the court should not place the applicant for the foreseeable future in a stronger economic position than the respondent. Whilst this sharing of risk and loss sounds very egalitarian, it might well have meant that where there was an unequal distribution of assets between cohabitants, an economically weaker primary carer partner would not on separation have been able to recover their full loss from a very wealthy former partner. Equal sharing of loss would not always have been fair. The attractiveness of the scheme is its promotion of financial autonomy between the partners. Indeed, some respondents to Barlow et al’s most recent BSA (see Table 14.1) and follow-up study recognised the Law Commission’s approach to be appropriate for married and unmarried childless couples. As Noreen, a cohabitant in her thirties with one child, explaining her view on the position in the scenario where a childless couple breaks up after a two-year marriage or cohabitation, confirmed: … there’s no children involved, why should one person be responsible for supporting you financially? However, the fact that the Law Commission scheme deliberately did not directly address the financial needs of those facing separation, and so would have been likely to have operated harshly on those who were post-separation primary carers of children of the relationship in some cases was not universally welcomed. The Law Commission took the view that relationship-generated economic disadvantage was a fairer and more specific way of defining and thereby addressing (albeit indirectly) the sort of post-cohabitation relationship need worthy of legal redress where, unlike

316 Anne Barlow in marriage, the parties had not publicly declared their commitment to the relationship (Law Commission, 2007: Appendix C, [C9]–[C12]). Indeed the BSA Survey did reveal some support for this approach (see Table 14.1 above) as did the followup study. As Ruth, in her thirties and a cohabitant with three children, saw it: [I]n a partnership people make compromises and quite often it’s the woman that makes compromises and that’s part of the deal. If it then turns sour … they shouldn’t be penalised because they’ve made sacrifices. Where there were children, though, the appetite for a difference in approach between married and cohabiting families was greatly reduced, and evidence of thinking that there should be a functional approach to families with children based on their needs is found across the recent studies: I don’t think it [legal protection] should be automatically tied into marriage or cohabiting or just living separately. It should be to do with the rights everybody has to have a minimum safety net. It should be to do with the rights that children have to have their parents have an income with which to bring up the children, regardless of, you know, whether the parents are still in a relationship or not. (Daniel, cohabitant of nine years, two children, respondent in Barlow et al’s 2007 study) Similarly, in Cooke, Barlow and Callus’s 2006 study, when asked for reasons why a cohabiting mother with children should have the same rights to the family home owned by her partner as a married mother on relationship breakdown, a married respondent explained: I think she should be allowed to stay in the house and then the property sold. . . . [Why?] . . . Because they’ve still got the same responsibilities to each other and their children. (Cooke, Barlow and Callus, 2006: 32) However, probably the biggest problems with the Law Commission’s scheme were its apparent complexity and its unfamiliar approach which resulted in being received less enthusiastically than it in fact deserved. Despite the Law Commission’s misgivings about equal treatment of functionally similar married and cohabiting couples, research does now clearly show that there is public (if not political) support for reconsideration of the hierarchy of those whom the law provides may obtain financial provision on relationship breakdown. As has been seen, there is greater support for financial provision after a two-year relationship for a cohabitant who has a child than there is for a spouse after a twoyear childless marriage. It is not clear to people why Mrs Miller should have got £5 million after a two-year childless marriage to a very rich man (Miller; McFarlane10), whereas Mrs Burns, a cohabitant of 19 years’ standing with another rich man, who brought up two children and worked part time, received nothing (Burns v Burns11). 10 11

[2006] UKHL 24. [1984] Ch 317.

Legal Rationality and Family Property 317 If we can agree that it is couples with children who are the most deserving of the law’s remedies, as public opinion seems to confirm, then we need to reformulate the hierarchy to reflect society’s consensus on this point, which is one with which, outside the sphere of family property, family law agrees. At the very least, it should be possible to extend to cohabitants the provision of reasonable requirements—a needs-based approach to financial provision on separation—as existed for married couples before the cases of White v White12 and Miller; McFarlane introduced the principle of equal sharing of matrimonial assets on divorce in higher asset cases under the Matrimonial Causes Act 1973. The Law Commission rejected its consultees’ call for a needs-based redistribution of assets on relationship breakdown as it felt there was not a clear, principled justification for needs-based relief that would help to determine how ‘need’ should be measured, which ‘needs’, if any, should be met by a partner, in what circumstances and for how long’ (Law Commission, 2007: Appendix C, [C9]).

However, these are all matters which the courts have had to address in the divorce context. In the majority of cases where assets do not exceed needs, it is the housing needs of the children and then those of the primary carer which are addressed first, followed by the housing needs of the other parent. The future income needs of all would then be considered and addressed. Before White and Miller; McFarlane, the other assets were left undisturbed (see Bird, 2000) but would now be subject to the equal sharing principle if they were classified as matrimonial assets in the divorce context. The courts are familiar with exercising their discretion in this context and could, as they used to on divorce, generously interpret needs in higher asset cases. Thus implementation of such a scheme would, from a practical point of view, be relatively easy and address the concerns of the public without wading into the difficulties which the equal sharing principle poses in the non-contractual cohabitation context (see Law Commission, 2007, Appendix C, [C5]–[C8]). Where there are no children, on the other hand, then financial autonomy may be a better goal for both the married and unmarried cohabitants and would encompass the call for enforceable pre-marital and cohabitation agreements, alongside a safety net of legal redress based on relationship-generated economic disadvantage and retained benefit. CONCLUSION

Love then has everything to do with it, as it makes law’s subjects far less susceptible to legally rational behaviour, within the current legal framework, in the context of their private lives. This means that we follow our hearts, often with the best of motives, and social norms endorse this approach. People do what is right for them in the context of their own lives and to act legally rationally, given the demands 12

[2001] 1 AC 596.

318 Anne Barlow of the current law, is often seen as inappropriate or too difficult. To ignore the social trends away from marriage leaves law looking rather inadequate in the face of social change. It is failing to protect the most vulnerable ‘uneven’ partners and it is failing to recognise that there are many ways to ‘do family’ (Morgan, 2000) which serve society well and which are worthy of support when things go wrong. Certainly, in the public mind, there is an equality discourse between functionally similar couples—where there are children and where there is ‘marriage-like’ behaviour which is borne out of a freedom to choose which partnering and parenting structure to adopt according to the context of your own life, perhaps best understood as a social interpretation of the meaning of one’s right to private and family life free from state interference. Indeed, this idea fits with some legal dicta—Hale LJ (as she then was) in SRJ v DWJ (Financial Provision)13 observed: It is not only in [the child’s] interests but in the community’s interests that parents, whether mothers or fathers and spouses, whether husbands or wives, should have a real choice between concentrating on breadwinning and concentrating on home-making and child-rearing, and do not feel forced, for fear of what might happen should their marriage break down much later in life, to abandon looking after the home and the family to other people for the sake of maintaining a career.

To find a way forward, we need to consider regulation of family property disputes across both married and cohabitation contexts in the light of twenty-first century values. The only clear alternative is to allow chaos to continue to reign, rather than socially locate family law within the real-life situations found in society. The best way forward, in my view, involves coming back to Teubner’s solution of reflexive law (Teubner, 1993), which, as King and Piper (1995) explain: is offered as an answer to the crisis caused by the failure of legal rationality under modern conditions to provide law with the necessary tools to restore consensual, moral and political values.

This approach offers a ‘bottom-up’ solution, which departs from the normal uniform standard-setting role that law has traditionally played and would permit a pluralistic approach to family regulation. That is not to say it would abandon standard-setting altogether. Rather that it would attempt to meet the needs of a society where partnerships founded on love take diverse forms, by being better attuned to shifting public attitudes and respecting the need for greater flexibility for couples to make their own arrangements whilst acknowledging minimum standards. Family property is surely a good starting point to examine how this could be done and protection for weaker economic partners where there are children, and regardless of whether or not they happen to have married, is clearly publicly acceptable, if not politically so. There is also a taste for private ordering where

13

[1999] 2 FLR 179 at 182.

Legal Rationality and Family Property 319 there are no children or where a couple have carefully thought through the implications of their separation for themselves and their whole family. Pre-marital and cohabitation agreements must surely have their day and allow couples to separate their financial and emotional layers of commitment where they are both fully cogniscent of their choices. By combining these features of autonomy and protection, we should at lest be en route to achieving a new normal order of family law as advocated by Henaghan (2008) and Eekelaar (2006) which is firmly socially located in current family practices and public expectations and yet is flexible enough to empower rather than enslave future generations. BIBLIOGRAPHY Atkin, B (2001) ‘Reforming property division in New Zealand: from marriage to relationships’ 3 European Journal of Law Reform 349. —— (2003) ‘The challenge of unmarried cohabitation—the New Zealand response’ 37 Family Law Quarterly 303. Barlow, A, Burgoyne, C, Clery, E and Smithson, J (2008a) ‘Cohabitation and the law: myths, money and the media’ in A Park, J Curtice, K Thompson, M Phillips and E Clery (eds) British Social Attitudes: The 24th Report (London, Sage). —— (2008b) Cohabitation and the Law: Myths, Money and the Media: Key Findings (London, Nuffield Foundation). Barlow, A, Burgoyne, C and Smithson, J (2007) The Living Together Campaign—An investigation of its impact on legally aware cohabitants (London, Ministry of Justice), available at http://www.justice.gov.uk/publications/research250707.htm. Barlow, A, Duncan, S, James, G and Park, A (2001) ‘Just a piece of paper? Marriage and cohabitation in Britain’ in A Park, J Curtice, K Thomson, L Jarvis and C Bromley (eds) British Social Attitudes: The 18th Report (London, Sage). —— (2005) Cohabitation, Marriage and the Law: Social Change and Legal Reform in the 21st Century (Oxford, Hart Publishing). Barlow, A and Duncan, S (2000) ‘Family law, moral rationalities and New Labour’s communitarianism: Part II’ 22 Journal of Social Welfare and Family Law 129. Beck, U and Beck-Gernsheim, E (1995) The Normal Chaos of Love (Cambridge, Polity Press). Bird, R (2000) ‘Ancillary relief outcomes’ 30 Family Law 831. Cooke, E, Barlow, A and Callus, T (2006) Community of Property: A Regime for England and Wales? (Bristol, The Policy Press). Dewar, J (1998) ‘The normal chaos of family law’ 61 Modern Law Review 467. Department of Trade and Industry, Women and Equality Unit (2003) Civil Partnership: A Framework for the Legal Recognition of Same-sex Couples (London, Department of Trade and Industry). Deech, R (1980) ‘The case against the legal recognition of cohabitation’ 29 International and Comparative Law Quarterly 480. Dworkin, R (1978) Taking Rights Seriously (Cambridge, MA, Harvard University Press). Eekelaar, J (2006) Family Law and Personal Life (Oxford, Oxford University Press). Freeman, M (1984) ‘Legal ideologies, patriarchal precedents and domestic violence’ in M Freeman (ed), State Law and Family: Critical Perspectives (London, Tavistock). —— (1985) ‘Towards a critical theory of family law’ 38 Current Legal Problems 153.

320 Anne Barlow Graycar, R and Millbank, J (2007) ‘From functional family to spinster sisters: Australia’s distinctive path to relationship recognition’ 24 Journal of Law and Policy 121. Henaghan, M (2008) ‘The Normal Order of Family Law’ 28 Oxford Journal of Legal Studies 165. Hibbs, M, Barton, C and Beswisk, J (2001) ‘Why marry? Perceptions of the affianced’ (2001) 31 Family Law 197. Home Office (1998) Supporting Families: A Consultative Document (London, TSO). Kiernan, K, Barlow, A and Merlo, R (2007) ‘Cohabitation Law Reform and its Impact on Marriage: Evidence from Australia and Europe’ 63 International Family Law 71. King, M and Piper, C (1995) How the Law Thinks about Children (Aldershot, Arena). Law Commission (2006) Cohabitation: The Financial Consequences of Relationship Breakdown—A Consultation Paper, Law Com CP No 179 (London, TSO). —— (2007) Cohabitation: The Financial Consequences of Relationship Breakdown, Law Com No 307 (London, TSO). Lewis, J (2001) The End of Marriage? Individualism and Intimate Relationships. (Cheltenham, Edward Elgar). Maclean, M and Eekelaar, J (2004) ‘Marriage and the moral bases of personal relationships’ 31 Journal of Law and Society 510. Morgan, D (1999) ‘Risk and family practices: accounting for change and fluidity in family life’ in S Smart and E Silva (eds), The New Family (London, Sage). Morgan, P (2000) Marriage-lite: The Rise of Cohabitation and its Consequences (London, Institute for the Study of Civil Society). National Statistics Online (2008) Focus on gender: living arrangements London, available at www.statistics.gov.uk/CCI/nugget.asp?ID=1652, accessed 31 October 2008. Probert, R (2006) ‘Cohabitation: contributions and sacrifices’ 36 Family Law 1060. Smart, C and Stevens, P (2000) Cohabitation Breakdown (London: FPSC). Social Justice Policy Group (2007) Breakthrough Britain, Volume 1: Family Breakdown, available at http://povertydebate.typepad.com/home/files/family_breakdown.pdf. Tennant, R, Taylor, J and Lewis, J (2006) Separating from Cohabitation: Making Arrangements for Finances and Parenting (London, Department for Constitutional Affairs), available at http://www.dca.gov.uk/research/2006/07_2006.pdf. Teubner, G (1993) ‘Substantive and reflexive elements in modern law’ Law and Society Review 239.

Index Aassve, A 230 Akerlof, GA 299 alternatives to litigation background 207–9 collaborative law 213–14 court order conversion 215–16 DIY settlements 211–12 mediation 212–13 pre/post-nuptial agreements 209–10 solicitor negotiation 214–15 summary 209 walking away 210–11 see also court proceedings ancillary relief 6–8, 185–205 background 185–6 conclusion 204–5 distinction between cases 194–5 judicial/court variations 198–205 low/no-money case 192–4 middle-money case 187–91 variations 191–2 non-legal considerations 195–8 reform proposals 221 research methodology 186–7 Arber, S 260 Ashby, K 144, 154 Avellar, S 247 Barlow, A 150, 182, 305, 309, 315, 316 Beck, U 305 Beck-Gernsheim, E 305 Blumstein, P 79 borrowing see consumer borrowing Bouman, AM 247 Bridge, S 289 Burgoyne, CB 83, 105, 144, 150, 154 Burns v Burns 311, 316 Calabresi, G 289 Callus, T 316 Campaign for Justice in Divorce (CJD) 27, 28 Child Maintenance and Other Payments Act 2008 10, 29–30, 224 child support 169, 224 Child Support Act 1991 10, 29–30, 31 children disparity of outcomes 25–6 financial impacts 125–7 financial provision for 10–11

focus on 29–30 separating cohabitants and 151 Children Act 1989 177, 178, 224 civil partnerships 304, 313 CJD (Campaign for Justice in Divorce) 27, 28 Coase theorem 295–7 cohabitation behaviour changes 297 categories 150 control/independence 99–100 death, attitudes to 310–11 hierarchy reconsideration 314–17 inconsistencies 306–7 intentions 89–90, 103–5 joint pool 96–7 legal position 5–6, 30–1 money management issues 62–3, 83–4, 95–6 typology 164–6 ownership see perceived ownership partial pooling 97–8 pragmaticism 98–9 public information campaign 306 research 90–2, 92–4, 92–3, 94–5 separate accounts 98–100 separation see separating cohabitants subcategories 68–70, 73–8 trends 297 typology 95–6, 309 as unregulated area 207–300 see also civil partnerships; marriage Cohen, L 292–4, 297 collaborative law 213–14 Collins, I 186 common law marriage myth 306–7, 307–8, 309 compensation 175–6 constructive trusts 139 consumer borrowing 110–14 background 110–11 correlates 111–13 reasons for 113–14 saving and 117–18 see also over-indebtedness Cooke, E 316 court order conversion 215–16 court proceedings children and 224 consistency 220–1 final hearing 220

322 Index financial dispute resolution hearing 218–20 first directions appointment 218, 222 Pre-Action Protocol 216–17 procedure 217–29 reluctant/obstructive litigants 222–4 summary 224–5 see also alternatives to litigation Cowan v Cowan 204–5 credit use see consumer borrowing Cretney, S 186 Davis, G 186 dependency 165–6, 169, 180–2, 183 ratio 41 Dewar, J 303, 305, 312 DiPrete, TA 230 divorce 227–54 ancillary relief 6–8 background 227–9 changes in effects over time 244 changing objectives 26–7 compensation for costs 289–90 conclusions 254–5 consumption based allocations 252–4 data use 230–3 divorcees’ characteristics 234–5 empirical research 28–9, 34–7 extent of loss 235–9 financial disputes research 34–7 income changes 240–2 income changes over time 236–8 long-term support 31–3 potential earnings 248–52 public debate 27–9 recovery 238–9, 242–4 regression analysis 239–47 repartnering rates 239 separation from marriage 231, 232, 233–9, 246 sources of data 229–30 women with high education 235–6 income growth 240, 241, 244, 245 Table with low education 235–6, 238–9 recovery 238–9, 242–4 Divorce Reform Act 1969 26 DIY settlements 211–12 Dnes, AW 252 domestic violence, financial difficulties and 131 Duncan, GJ 229 economic analysis background 285 Coase theorem 295–7 cohabitation studies 297–300

enlightened thought 285–6 family law questions 300–1 life profile theory 292–5, 297 normative analysis 286–90, 291 positivist analysis 291–2 education divorce see under divorce, women pension accumulation 265–8 Eekelaar, J 33, 38, 309, 311, 319 Ellman, I 36 empirical data accessibility 14 importance 3–4 interpretation 14–15 normative judgments and 12–13 role 12–15 structured discretion and 14 enlightened thought 285–6 Estaugh, V 70, 149 Evason, E 258 Family Bar 33, 34–7 Family Law (Scotland) Act 1985 221 Family Law Act 1996 26 family law reform background 25–6 changing focus 29–30 differentiation to inclusiveness 30–1 pragmatism and empiricism 26–9 qualitative research 34–7 see also economic analysis Family Proceedings Rules 1991 217–18, 224 family raising 125–7 Fatal Accidents Act 1976 306 FDA (first directions appointment) 218, 222, 223 FDR see financial dispute resolution financial difficulties, life course relationships and background 109–10, 122, 132 breakdown see relationship breakdown, financial impacts children/family raising 125–7, 131–2 partnering 123–5 see also consumer borrowing; overindebtedness; saving financial dispute resolution (FDR) 197–8 hearing 218–20 first directions appointment (FDA) 218, 222, 223 Fowler v Barron 9 gender patterns 129–30 work balance see paid/unpaid work balance Gershuny, J 50 Giddens, A 79, 84

Index 323 Ginn, J 260 Gissing v Gissing 287 graduate earnings 45 Gupta, S 229

litigation friend 223 see also alternatives to litigation; court proceedings Living Together Campaign 306, 308–9, 309

Hakim, C 54 Hale, B 31, 140–1, 150 happiness, and irrational behaviour Haskey, J 149 Henaghan, M 319 Hoffman, SD housing tenure 112

McKay, S 278 McKeever, M 229, 244 Maclean, M 309 McManus, PA 230 male-breadwinner model 50 Manning, WD 229 Manting, D 247 marriage Coase theorem 295–7 encouragement of 313 financial impacts 124–5 hierarchy reconsideration 314–17 life profile theory 292–5 life-profile model 297 money management issues 61–2, 83–4 subcategories 68–9, 73–8 pooling system and 90–1 see also cohabition; spouses Matrimonial Causes Act 1973 26, 29, 30, 31, 34, 185, 198, 207, 221, 289, 317 Matrimonial and Family Proceedings Act 1984 6 Mechoulan, S 291, 295, 298 mediation 212–13 Melamed, AD 289 Miller; McFarlane 7, 12, 15, 32, 185–6, 221, 229, 251–2, 254, 255, 289–90, 311–12, 316–17 money management patterns background 61–3 changes 66–7 cohabiting/married categories 68–70, 73–8 conclusion 83–4 data in analysis 67–8 independent management system 65–6 individualised patterns 78–81 satisfaction/happiness findings 81–3 socio-economic characteristics 70–3 as taboo subject 64 typology 64–5 motherhood effect 263–5

286

inference, and imputation 286–8 Inheritance (Provision for Family and Dependants) Act 1975 306 International Social Survey Programme (ISSP) 63 irrational behaviour 19 n, 286 see also legal rationality ISSP (International Social Survey Programme) 63 James v Thomas 9 Jarvis, S 230 Jenkins, SP 230, 244 joint/separate bank accounts

91, 105

Kaldor-Hicks test 290 Katz, ML 299 Kiernan, KE 70, 149 labour market gender balance see paid/unpaid work balance Laskar v Laskar 149 Law Commission 11, 14, 180–2, 314–17 legal aid 215 legal context 5–6 ancillary relief 6–8 asset division 8–10 financial provision for children 10–11 reform proposals 11–12 research see empirical data legal rationality 303–19 assumption 304–5 avoidance of mistake 306–7 background 303–4 conclusion 317–19 current practice rationale 311–13 hierarchy reconsideration 314–17 love-based relationships and 305 reflexive law 318–19 response issues 307–11 Lewis, J 50, 73, 309 life course relationships see financial difficulties, life course relationships and life profile theory 292–5, 297

negotiation by solicitor 214–15 normative analysis 286–90, 291 Nozick, R 286 nuisance cases 295–6 over-indebtedness background 118–19 correlates 119–20 reasons for 120–2 see also consumer borrowing ownership see perceived ownership

324 Index Pahl, J 64, 66, 73, 90, 105 paid/unpaid work balance background 41–2 compulsion 55–6 female participation 42–4 housework patterns 47–50 issues 55–6 maternal employment 47, 48 Table mobility after childbirth 43–4 pay equality 44–6 policy suggestions 55–7 wage equalisation 56–7 within couples 46–50 work-family see work-family balance Parkman, A 289, 290 partnering 123–5 Peggs, K 258 pension accumulation attitudes to 168 background 257–60 conclusions 278–9 data/methodology 261 inequalities 259–60 earnings 262 education and 265–8 motherhood effect 263–5 partnership dependency 268–72 status 272–8 questions 258–9 third tier provision 260–1, 272–4, 275 Table, 279 perceived ownership 100–3, 104–5 background 100–1 blurred ownership 102 distinct ownership 101–2, 103 Table shared ownership 102–3 Pettitt v Pettitt 287 positivist analysis 291–2 Pre-Action Protocol 216–17 pre/post-nuptial agreements 209–10 Probert, R 141 property cohabitants see separating cohabitants, property issues and trusts law 8–10 rational behaviour 19 n, 286 see also legal rationality Rawls, J 286 reflexive law 318–19 relationship breakdown, financial impacts 127–32 background 127–8 domestic violence and 131 gender patterns 129–30 outcomes effect 130–1 as reason for difficulties 129

timing effect 128–9 see also divorce; separating cohabitants Rogers, S 73 Rowlingson, K 278 Rowthorn, R 299 saving background 114–15 borrowing and 117–18 correlates 115–16 reasons for 116–17 Schwarzt, P 79 separating cohabitants asset division 8–10 attitudes to 310–11, 312 differentiation to inclusiveness 30–1 disadvantage following background 174 compensation 175–6 dependency 165–6, 169, 180–2, 183 Law Commission recommendations 180–2 legal entitlement 176–9 material circumstances 174–5 multiple forms 179 financial arrangements background 161–2 child support 169 contributions issues 171 debt 168 during relationship 164–6 emotional/psychological influences 173–4 equality issues 173 financial assets 167–8 homes 167 household assets 168–9 legal entitlement 173 needs issues 172 ownership issues 169–71, 182 rights/responsibilities conferred 182–3 study 162–4 financial disputes research 34–7 property issues 139–59 advice at time of purchase 144–5 arrangement of finances 154–6 background 139 capital moneys receipt 147–8 dependant children 151 discharge of outgoings 156–7 discussions at time of transfer 142–3 individual characters/personalities 157–8 initial financial contributions 151–2 intention factors 140–1, 158–9 joint names acquisition 145–7 nature of relationship 149–51 non-financial contributions 153–4 purpose of acquisition 148–9 recollection of advice 143–4

Index 325 study 141–2 subsequent financial contributions 152–3 reflexive law and 318–19 reform proposals 314–17 research needs 33–4 retained benefits 314–15 see also relationship breakdown SFLA (Solicitor Family Law Association) 208 Smart, C 156, 309 Smithson, J 150 Smock, PJ 229, 247 socio-legal research see empirical data Solicitor Family Law Association (SFLA) 208 solicitor negotiation 214–15 Spence, L 258 spouses legal position 5–6, 30–1 long-term support 31–3 SRJ v DWJ (Financial Provisions) 318 Stack v Dowden 12, 14, 15, 16, 89, 139–41, 146–52 passim, 158, 233, 287–8, 311 Stevens, P 156, 309 Sturges v Bridgman 296 Teubner, G 318 Thatcher, Margaret 29 Trebilcock, M 289

trusts constructive 139 property and 8–10 utilitarianism 286 Uunk, W 230 walking away 210–11 Warren, T 260 Weitzman, L 28 welfare economics 286 White, L 73 White v White 7, 12, 32, 185, 221, 312 Wolfinger, NH 229, 244 women see under divorce work-family balance 50–5 background 50 male-breadwinner model 50 parental leave 53–4 range of models 50–2 redistribution 54–5 time-use patterns 52–3 Working Time Directive (EU) 53 Yellen, JL 299 Young, Baroness Zelder, M 296

304–5